kth.sePublications
Change search
Refine search result
123 1 - 50 of 132
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Rows per page
  • 5
  • 10
  • 20
  • 50
  • 100
  • 250
Sort
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
  • Disputation date (earliest first)
  • Disputation date (latest first)
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
  • Disputation date (earliest first)
  • Disputation date (latest first)
Select
The maximal number of hits you can export is 250. When you want to export more records please use the Create feeds function.
  • 1. Basit, K. A.
    et al.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    GUMO inspired ontology to support user experience based Citywide Mobile Learning2011In: Proc. - Int. Conf. User Sci. Eng., i-USEr, 2011, p. 195-200Conference paper (Refereed)
    Abstract [en]

    User experience has been extensively discussed in literature, yet the idea of applying it to explain and comprehend the conceptualization of Mobile Learning (ML) is relatively new. Consequently much of the existing works are mainly theoretical and they concentrate to establish and explain the relationship between ML and experience. Little has been done to apply or adopt it into practice. In contrast to the currently existing approaches, this paper presents an ontology to support Citywide Mobile Learning (CML). The ontology presented in this paper addresses three fundamental aspects of CML, namely User Model, User Experience and Places/Spaces which exist in the city. The ontology presented here not only attempts to model and translate the theoretical concepts such as user experience and Place/Spaces for citywide context for Mobile Learning, but also apply them into practice. The discussed ontology is used in our system to support Place/Space based CML.

  • 2.
    Bunea, Ramona
    et al.
    KTH, School of Information and Communication Technology (ICT).
    Mokarizadeh, Shahab
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Dokoohaki, Nima
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Exploiting dynamic privacy in socially regularized recommenders2012In: Data Mining Workshops (ICDMW), 2012 IEEE 12th International Conference on, IEEE , 2012, p. 539-546Conference paper (Refereed)
    Abstract [en]

    In this paper we introduce a privacy-aware collaborative filtering recommender framework which aims to address the privacy concern of profile owners in the context of social trust sparsity. While sparsity in social trust is mitigated by similarity driven trust using a probabilistic matrix factorization technique, the privacy issue is addressed by employing a dynamic privacy inference model. The privacy inference model exploits the underlying inter-entity trust information to obtain a personalized privacy view for each individual in the social network. We evaluate the proposed framework by employing an off-the-shelf collaborative filtering recommender method to make predictions using this personalized view. Experimental results show that our method offers better performance than similar non-privacy aware approaches, while at the same time meeting user privacy concerns.

  • 3. Cena, Federica
    et al.
    Dokoohaki, Nima
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Electronic, Computer and Software Systems, ECS.
    Forging Trust and Privacy with User Modeling Frameworks: An Ontological Analysis2011In: The First International Conference on Social Eco-Informatics: (SOTICS 2011) / [ed] Dokoohaki and Hall, IARIA , 2011, p. 43-48Conference paper (Refereed)
    Abstract [en]

    With the ever increasing importance of social net- working sites and services, socially intelligent agents who are responsible for gathering, managing and maintaining knowledge surrounding individual users are of increasing interest to both computing research communities as well as industries. For these agents to be able to fully capture and manage the knowledge about a user’s interaction with these social sites and services, a social user model needs to be introduced. A social user model is defined as a generic user model (model capable of capturing generic information related to a user), plus social dimensions of users (models capturing social aspects of user such as activities and social contexts). While existing models capture a proportion of such information, they fail to model and present ones of the most important dimensions of social connectivity: trust and privacy. To this end, in this paper, we introduce an ontological model of social user, composed by a generic user model component, which imports existing well-known user model structures, a social model, which contains social dimensions, and trust, reputation and privacy become the pivotal concepts gluing the whole ontological knowledge models together.

    Download full text (pdf)
    cena-dokoohaki-sotics2011
  • 4. Chang, Carl K.
    et al.
    Gao, YanHurson, AliMatskin, MihhailKTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.McMillin, BruceOkabe, YasuoSeceleanu, CristinaYoshida, Kenichi
    IEEE 38th Annual International Computers, Software and Applications Conference: Västerås, Sweden - July 21-25, 20142014Conference proceedings (editor) (Refereed)
  • 5. Claycomb, W.
    et al.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Nakamura, M.
    Message from the Workshop Chairs - Part III2015In: Proceedings - International Computer Software and Applications Conference, IEEE Communications Society, 2015, Vol. 3Conference paper (Refereed)
  • 6. Corodescu, A. -A
    et al.
    Nikolov, N.
    Khan, A. Q.
    Soylu, A.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Payberah, Amir H.
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Roman, D.
    Locality-aware workflow orchestration for big data2021In: ACM International Conference Proceeding Series, Association for Computing Machinery , 2021, p. 62-70Conference paper (Refereed)
    Abstract [en]

    The development of the Edge computing paradigm shifts data processing from centralised infrastructures to heterogeneous and geographically distributed infrastructure. Such a paradigm requires data processing solutions that consider data locality in order to reduce the performance penalties from data transfers between remote (in network terms) data centres. However, existing Big Data processing solutions have limited support for handling data locality and are inefficient in processing small and frequent events specific to Edge environments. This paper proposes a novel architecture and a proof-of-concept implementation for software container-centric Big Data workflow orchestration that puts data locality at the forefront. Our solution considers any available data locality information by default, leverages long-lived containers to execute workflow steps, and handles the interaction with different data sources through containers. We compare our system with Argo workflow and show significant performance improvements in terms of speed of execution for processing units of data using our data locality aware Big Data workflow approach. 

  • 7.
    Corodescu, Andrei-Alin
    et al.
    Univ Oslo, Dept Informat, N-0373 Oslo, Norway..
    Nikolov, Nikolay
    SINTEF AS, Software & Serv Innovat, N-0373 Oslo, Norway..
    Khan, Akif Quddus
    Norwegian Univ Sci & Technol, Dept Comp Sci, N-2815 Gjovik, Norway..
    Soylu, Ahmet
    OsloMet Oslo Metropolitan Univ, Dept Comp Sci, N-0166 Oslo, Norway..
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Payberah, Amir H.
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Roman, Dumitru
    SINTEF AS, Software & Serv Innovat, N-0373 Oslo, Norway..
    Big Data Workflows: Locality-Aware Orchestration Using Software Containers2021In: Sensors, E-ISSN 1424-8220, Vol. 21, no 24, article id 8212Article in journal (Refereed)
    Abstract [en]

    The emergence of the edge computing paradigm has shifted data processing from centralised infrastructures to heterogeneous and geographically distributed infrastructures. Therefore, data processing solutions must consider data locality to reduce the performance penalties from data transfers among remote data centres. Existing big data processing solutions provide limited support for handling data locality and are inefficient in processing small and frequent events specific to the edge environments. This article proposes a novel architecture and a proof-of-concept implementation for software container-centric big data workflow orchestration that puts data locality at the forefront. The proposed solution considers the available data locality information, leverages long-lived containers to execute workflow steps, and handles the interaction with different data sources through containers. We compare the proposed solution with Argo workflows and demonstrate a significant performance improvement in the execution speed for processing the same data units. Finally, we carry out experiments with the proposed solution under different configurations and analyze individual aspects affecting the performance of the overall solution.

  • 8.
    Dautaras, Justas
    et al.
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Mobile Crowdsensing with Imagery Tasks2021In: Proceedings 2021 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology, Association for Computing Machinery (ACM) , 2021, p. 54-61Conference paper (Refereed)
    Abstract [en]

    The amount of gadgets connected to the internet has grown rapidly in the recent years. These human owned devices can potentially be used to gather sensor data without active involvement of their owners. One of the types of platforms that contribute to the utilisation of these devices are mobile crowdsensing systems. These systems can be used for different tasks including different types of community support. While these systems are quite widely used, yet little research has been done for integration of imagery data into them which require also human involvement. This paper considers a mobile crowdsensing system where gathering data from sensors is supported by crowdsourcing human intelligence for providing both textual and visual information. We also explore the best settings for such a system. Imagery processing is integrated into an already existing mobile crowdsensing platform CrowdS. The solution was evaluated both by a limited number of real life users and by conducting simulations. The simulations represent complex scenarios with multi-level variables. The results of simulation allow suggest an efficient configuration for the parameters and characteristics of the environment used in imagery integration.

  • 9.
    Dessalk, Yared Dejene
    et al.
    KTH.
    Nikolov, N.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Soylu, A.
    Roman, D.
    Scalable Execution of Big Data Workflows using Software Containers2020In: Proceedings of the 12th International Conference on Management of Digital EcoSystems, MEDES 2020, Association for Computing Machinery, Inc , 2020, p. 76-83Conference paper (Refereed)
    Abstract [en]

    Big Data processing involves handling large and complex data sets, incorporating different tools and frameworks as well as other processes that help organisations make sense of their data collected from various sources. This set of operations, referred to as Big Data workflows, require taking advantage of the elasticity of cloud infrastructures for scalability. In this paper, we present the design and prototype implementation of a Big Data workflow approach based on the use of software container technologies and message-oriented middleware (MOM) to enable highly scalable workflow execution. The approach is demonstrated in a use case together with a set of experiments that demonstrate the practical applicability of the proposed approach for the scalable execution of Big Data workflows. Furthermore, we present a scalability comparison of our proposed approach with that of Argo Workflows-one of the most prominent tools in the area of Big Data workflows.

  • 10.
    Dokoohaki, Nima
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Kaleli, Cihan
    Polat, Huseyin
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Achieving Optimal Privacy in Trust-Aware Social Recommender Systems2010In: SOCIAL INFORMATICS / [ed] Bolc L; Makowski M; Wierzbicki A, 2010, Vol. 6430, p. 62-79Conference paper (Refereed)
    Abstract [en]

    Collaborative filtering (CF) recommenders are subject to numerous shortcomings such as centralized processing, vulnerability to shilling attacks, and most important of all privacy. To overcome these obstacles, researchers proposed for utilization of interpersonal trust between users, to alleviate many of these crucial shortcomings. Till now, attention has been mainly paid to strong points about trust-aware recommenders such as alleviating profile sparsity or calculation cost efficiency, while least attention has been paid on investigating the notion of privacy surrounding the disclosure of individual ratings and most importantly protection of trust computation across social networks forming the backbone of these systems. To contribute to addressing problem of privacy in trust-aware recommenders, within this paper, first we introduce a framework for enabling privacy-preserving trust-aware recommendation generation. While trust mechanism aims at elevating recommenders accuracy, to preserve privacy, accuracy of the system needs to be decreased. Since within this context, privacy and accuracy are conflicting goals we show that a Pareto set can be found as an optimal setting for both privacy-preserving and trust-enabling mechanisms. We show that this Pareto set, when used as the configuration for measuring the accuracy of base collaborative filtering engine, yields an optimized tradeoff between conflicting goals of privacy and accuracy. We prove this concept along with applicability of our framework by experimenting with accuracy and privacy factors, and we show through experiment how such optimal set can be inferred.

  • 11.
    Dokoohaki, Nima
    et al.
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    An Adaptive Framework for Discovery andMining of User Profiles from Social Web-based Interest Communities2013In: The Influence of Technology on Social Network Analysis and Mining / [ed] Özyer, T.; Rokne, J.; Wagner, G.; Reuser, A., Wien: Springer, 2013, Vol. 23, p. 497-519Chapter in book (Refereed)
    Abstract [en]

    Abstract Within this paper we introduce an adaptive framework for semi- tofully-automatic discovery, acquisition and mining of topic style interest profilesfrom openly accessible social web communities. To do such, we build an adaptivetaxonomy search tree from target domain (domain towards which we are gatheringand processing profiles for), starting with generic concepts at root moving down tospecific-level instances at leaves, then we utilize one of proposed Quest schemesto read the concept labels from the tree and crawl the source social networkrepositories for profiles containing matching and related topics. Using machinelearning techniques, cached profiles are then mined in two consecutive steps,utilizing a clusterer and a classifier in order to assign and predict correct profilesto their corresponding clustered corpus, which are retrieved later on by an ontology-based recommender to suggest and recommend the community members with theitems of their similar interest. Focusing on increasingly important digital culturalheritage context, using a set of profiles acquired from an openly accessible socialnetwork, we test the accuracy and adaptivity of framework. We will show that a tradeoff between schemes proposed can lead to adaptive discovery of highly relevant profiles.

  • 12.
    Dokoohaki, Nima
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Effective Design of Trust Ontologies for Improvement in the Structure of Socio-Semantic Trust Networks2008In: International Journal On Advances in Intelligent Systems, ISSN 1942-2679, Vol. 1, no 1, p. 23-42Article in journal (Refereed)
  • 13.
    Dokoohaki, Nima
    et al.
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Mining divergent opinion trust networks through latent dirichlet allocation2012In: 2012 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, IEEE , 2012, p. 879-886Conference paper (Refereed)
    Abstract [en]

    While the focus of trust research has been mainly on defining and modeling various notions of social trust, less attention has been given to modeling opinion trust. When speaking of social trust mainly homophily (similarity) has been the most successful metric for learning trustworthy links, specially in social web applications such as collaborative filtering recommendation systems. While pure homophily such as Pearson coefficient correlation and its variations, have been favorable to finding taste distances between individuals based on their rated items, they are not necessarily useful in finding opinion distances between individuals discussing a trending topic, e. g. Arab spring. At the same time text mining techniques, such as vector-based techniques, are not capable of capturing important factors such as saliency or polarity which are possible with topical models for detecting, analyzing and suggesting aspects of people mentioning those tags or topics. Thus, in this paper we are proposing to model opinion distances using probabilistic information divergence as a metric for measuring the distances between people's opinion contributing to a discussion in a social network. To acquire feature sets from topics discussed in a discussion we use a very successful topic modeling technique, namely Latent Dirichlet Allocation (LDA). We use the distributions resulting to model topics for generating social networks of group and individual users. Using a Twitter dataset we show that learned graphs exhibit properties of real-world like networks.

  • 14.
    Dokoohaki, Nima
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Personalizing Human Interaction through Hybrid Ontological Profiling: Cultural Heritage Case Study2008In: 1st Workshop on Semantic Web Applications and Human Aspects, (SWAHA08), 2008, p. 133-140Conference paper (Refereed)
    Abstract [en]

    In this paper we present a novel user profile formalization, which allows describingthe user attributes as well as history of user access for personalized, adaptive and interactiveexperience while we believe that our approach is applicable to different semantic applicationswe illustrate our solution in the context of online and onsite museums and exhibits visit. Weargue that a generic structure will allow incorporation of multiple dimensions of user attributesand characteristics as well as allowing different abstraction levels for profile formalization andpresentations. In order to construct such profile structures we extend and enrich existingmetadata vocabularies for cultural heritage to contain keywords pertaining to usage attributesand user related keywords. By extending metadata vocabularies we allow improvedmatchmaking between extended user profile contents and cultural heritage contents. Thisextension creates the possibility of further personalization of access to cultural heritageavailable through online and onsite digital libraries.

  • 15.
    Dokoohaki, Nima
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    Norwegian University of Science and Technology.
    Quest: An Adaptive Framework for User Profile Acquisition from Social Communities of Interest2010In: Proceedings - 2010 International Conference on Advances in Social Network Analysis and Mining, ASONAM 2010, 2010, p. 360-364Conference paper (Refereed)
    Abstract [en]

    Within this paper we introduce a framework for semi- to full-automatic discovery and acquisition of bag-of-words style interest profiles from openly accessible Social Web communities. To do such, we construct a semantic taxonomy search tree from target domain (domain towards which we're acquiring profiles for), starting with generic concepts at root down to specific-level instances at leaves, then we utilize one of proposed Quest methods, namely Depth-based, N-Split and Greedy to read the concept labels from the tree and crawl the source Social Network for profiles containing corresponding topics. Cached profiles are then mined in a two-step approach, using a clusterer and a classifier to generate predictive model presenting weighted profiles, which are used later on by a semantic recommender to suggest and recommend the community members with the items of their similar interest.

  • 16.
    Dokoohaki, Nima
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Reasoning about Weighted Semantic User Profiles through Collective Confidence Analysis: A Fuzzy Evaluation2010In: ADVANCES IN INTELLIGENT WEB MASTERING-2, PROCEEDINGS    / [ed] Snasel V; Szczepaniak PS; Abraham A; Kacprzyk J, 2010, Vol. 67, p. 71-81Conference paper (Refereed)
    Abstract [en]

    User profiles are vastly utilized to alleviate the increasing problem of so called information overload. Many important issues of Semantic Web like trust, privacy, matching and ranking have a certain degree of vagueness and involve truth degrees that one requires to present and reason about. In this ground, profiles tend to be useful and allow incorporation of these uncertain attributes in the form of weights into profiled materials. In order to interpret and reason about these uncertain values, we have constructed a fuzzy confidence model, through which these values could be collectively analyzed and interpreted as collective experience confidence of users. We analyze this model within a scenario, comprising weighted user profiles of a semantically enabled cultural heritage knowledge platform. Initial simulation results have shown the benefits of our mechanism for alleviating problem of sparse and empty profiles.

  • 17.
    Dokoohaki, Nima
    et al.
    KTH, School of Information and Communication Technology (ICT), Microelectronics and Information Technology, IMIT.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Microelectronics and Information Technology, IMIT.
    Structural Determination of Ontology-Driven Trust Networks in Semantic Social Institutions and Ecosystems2007In: Proceedings of the International Conference on Mobile Ubiquitous Computing, Systems, Services and Technologies (UBICOMM'07) and the International Conference on Advances in Semantic Processing SEMAPRO, 2007, p. 263-268Conference paper (Refereed)
    Abstract [en]

    Social institutions and ecosystems are growing across the web and social trust networks formed within these systems create an extraordinary test-bed to study relation dependant notions such as trust, reputation and belief. In order to capture, model and represent the semantics of trust relationships forming the trust networks, main components of relationships are represented and described using ontologies. This paper investigates how effective design of trust ontologies can improve the structure of trust networks created and implemented within semantic web-driven social institutions and systems. Based on the context of our research, we represent a trust ontology that captures the semantics of the structure of trust networks based on the context of social institutions and ecosystems on semantic web.

  • 18.
    Dokoohaki, Nima
    et al.
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Afzal, Usman
    KTH, School of Information and Communication Technology (ICT).
    Islam, Md. Mistamikul
    KTH, School of Information and Communication Technology (ICT).
    An Enterprise Social Recommendation System for Connecting Swedish Professionals2014In: Proceedings - IEEE 38th Annual International Computers, Software and Applications Conference Workshops, COMPSACW 2014 / [ed] Carl K. Chang, Yan Gao, Ali Hurson, Mihhail Matskin, Bruce McMillin, Yasuo Okabe, IEEE Communications Society, 2014, p. 234-239Conference paper (Refereed)
    Abstract [en]

    Most cooperative businesses rely on some form of social networking system to facilitate user profiling and networking of their employees. To facilitate the discovery, matchmaking and networking among the co-workers across the enterprises social recommendation systems are often used. Off-the-shelf nature of these components often makes it hard for individuals to control their exposure as well as their preferences of whom to connect to. To this end, trust based recommenders have been amongst the most popular and demanding solutions due to their advantage of using social trust to generate more accurate suggestions for peers to connect to. They also allow individuals to control their exposure based on explicit trust levels. In this work we have proposed for an enterprise trust-based recommendation system with privacy controls. To generate accurate predictions, a local trust metric is defined between users based on correlations of user's profiled content such as blogging, articles wrote, comments, and likes along with profile information such as organization, region, interests or skills. Privacy metric is defined in such a way that users have full freedom either to hide their data from the recommender or customize their profiles to make them visible only to users with defined level of trustworthy.

  • 19. Dokoohaki, Nima
    et al.
    Zikou, Filippia
    KTH.
    Gillblad, Daniel
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Predicting Swedish Elections with Twitter: A Case for Stochastic Link Structure Analysis2015In: PROCEEDINGS OF THE 2015 IEEE/ACM INTERNATIONAL CONFERENCE ON ADVANCES IN SOCIAL NETWORKS ANALYSIS AND MINING (ASONAM 2015), Association for Computing Machinery (ACM), 2015, p. 1269-1276Conference paper (Refereed)
    Abstract [en]

    The question that whether Twitter data can be leveraged to forecast outcome of the elections has always been of great anticipation in the research community. Existing research focuses on leveraging content analysis for positivity or negativity analysis of the sentiments of opinions expressed. This is while, analysis of link structure features of social networks underlying the conversation involving politicians has been less looked. The intuition behind such study comes from the fact that density of conversations about parties along with their respective members, whether explicit or implicit, should reflect on their popularity. On the other hand, dynamism of interactions, can capture the inherent shift in popularity of accounts of politicians. Within this manuscript we present evidence of how a well-known link prediction algorithm, can reveal an authoritative structural link formation within which the popularity of the political accounts along with their neighbourhoods, shows strong correlation with the standing of electoral outcomes. As an evidence, the public time-lines of two electoral events from 2014 elections of Sweden on Twitter have been studied. By distinguishing between member and official party accounts, we report that even using a focus-crawled public dataset, structural link popularities bear strong statistical similarities with vote outcomes. In addition we report strong ranked dependence between standings of selected politicians and general election outcome, as well as for official party accounts and European election outcome.

  • 20.
    Erliksson, Karl Fredrik
    et al.
    KTH, School of Electrical Engineering and Computer Science (EECS), Electrical Engineering. Peltarion, Stockholm, Sweden.
    Arpteg, Anders
    Peltarion, Stockholm, Sweden.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Payberah, Amir H.
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Cross-Domain Transfer of Generative Explanations Using Text-to-Text Models2021In: Lecture Notes in Computer Science, Springer Nature , 2021, p. 76-89Conference paper (Refereed)
    Abstract [en]

    Deep learning models based on the Transformers architecture have achieved impressive state-of-the-art results and even surpassed human-level performance across various natural language processing tasks. However, these models remain opaque and hard to explain due to their vast complexity and size. This limits adoption in highly-regulated domains like medicine and finance, and often there is a lack of trust from non-expert end-users. In this paper, we show that by teaching a model to generate explanations alongside its predictions on a large annotated dataset, we can transfer this capability to a low-resource task in another domain. Our proposed three-step training procedure improves explanation quality by up to 7% and avoids sacrificing classification performance on the downstream task, while at the same time reducing the need for human annotations.

  • 21.
    Fazeli, Soude
    et al.
    KTH, School of Information and Communication Technology (ICT), Electronic, Computer and Software Systems, ECS.
    Zarghami, Alireza
    Dokoohaki, Nima
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Elevating Prediction Accuracy in Trust-aware Collaborative Filtering Recommenders through T-index Metric and TopTrustee lists2010In: Journal of Emerging Technologies in Web Intelligence, ISSN 1798-0461, Vol. 2, no 4, p. 300-309Article in journal (Refereed)
    Abstract [en]

    The growing popularity of Social Networks raises the important issue of trust. Among many systems which have realized the impact of trust, Recommender Systems have been the most influential ones. Collaborative Filtering Recommenders take advantage of trust relations between users for generating more accurate predictions. In this paper, we propose a semantic recommendation framework for creating trust relationships among all types of users with respect to different types of items, which are accessed by unique URI across heterogeneous networks and environments. We gradually build up the trust relationships between users based on the rating information from user profiles and item profiles to generate trust networks of users. For analyzing the formation of trust networks, we employ Tindex as an estimate of a user’s trustworthiness to identify and select neighbors in an effective manner. In this work, we utilize T-index to form the list of an item’s raters, called TopTrustee list for keeping the most reliable users who have already shown interest in the respective item. Thus, when a user rates an item, he/she is able to find users who can be trustworthy neighbors even though they might not be accessible within an upper bound of traversal path length. An empirical evaluation demonstrates how T-index improves the Trust Network structure by generating connections to more trustworthy users. We also show that exploiting Tindex results in better prediction accuracy and coverage of recommendations collected along few edges that connect users on a Social Network.

  • 22.
    Fazeli, Soude
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Zarghami, Alireza
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Dokoohaki, Nima
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Mechanizing Social Trust-Aware Recommenders with T-Index Augmented Trustworthiness2010In: TRUST, PRIVACY AND SECURITY IN DIGITAL BUSINESS  / [ed] Katsikas S; Lopez J; Soriano M, 2010, Vol. 6264, p. 202-213Conference paper (Refereed)
    Abstract [en]

    Social Networks have dominated growth and popularity of the Web to an extent which has never been witnessed before. Such popularity puts forward issue of trust to the participants of Social Networks. Collaborative Filtering Recommenders have been among many systems which have begun taking full advantage of Social Trust phenomena for generating more accurate predictions. For analyzing the evolution of constructed networks of trust, we utilize Collaborative Filtering enhanced with T-index as an estimate of a user's trustworthiness to identify and select neighbors in an effective manner. Our empirical evaluation demonstrates how T-index improves the Trust Network structure by generating connections to more trustworthy users. We also show that exploiting T-index results in better prediction accuracy and coverage of recommendations collected along few edges that connect users on a network.

  • 23.
    Fernando, Tharidu
    et al.
    KTH.
    Gureev, Nikita
    KTH.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Zwick, M.
    Natschlager, T.
    WorkflowDSL: Scalable Workflow Execution with Provenance for Data Analysis Applications2018In: Proceedings - International Computer Software and Applications Conference, IEEE Computer Society , 2018, p. 774-779Conference paper (Refereed)
    Abstract [en]

    Data analysis projects typically use different programming languages (from Python for prototyping to C++ for support of runtime constraints) at their different stages by different experts. This creates a need for a data processing framework that is re-usable across multiple programming languages and supports collaboration of experts. In this work, we discuss implementation of a framework which uses a Domain Specific Language (DSL), called WorkflowDSL, that enables domain experts to collaborate on fine-tuning workflows. The framework includes support for parallel execution without any specialized code. It also provides a provenance capturing framework that enables users to analyse past executions and retrieve complete lineage of any data item generated. Graph database is used for storing provenance data. Advantages of usage of a graph database compare to relational databases are demonstrated. Experiments which were performed using a real-world scientific workflow from the bioinformatics domain and industrial data analysis models show that users were able to execute workflows efficiently when using WorkflowDSL for workflow composition and Python for task implementations. Moreover, we show that capturing provenance data can be useful for analysing past workflow executions.

  • 24.
    Hammar, Kim
    et al.
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Jaradat, Shatha
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Dokoohaki, Nima
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Deep text classification of Instagram data using word embeddings and weak supervision2020In: WEB INTELLIGENCE, ISSN 2405-6456, Vol. 18, no 1, p. 53-67Article in journal (Refereed)
    Abstract [en]

    With the advent of social media, our online feeds increasingly consist of short, informal, and unstructured text. Instagram is one of the largest social media platforms, containing both text and images. However, most of the prior research on text processing in social media is focused on analyzing Twitter data, and little attention has been paid to text mining of Instagram data. Moreover, many text mining methods rely on training data annotated manually by humans, which in practice is both difficult and expensive to obtain. In this paper, we present methods for weakly supervised text classification of Instagram text. We analyze a corpora of Instagram posts from the fashion domain and train a deep clothing classifier with weak supervision to classify Instagram posts based on the associated text. With our experiments, we demonstrate that in absence of annotated training data, using weak supervision to train models is a viable approach. With weak supervision we were able to label a large dataset in hours, something that would have taken months to do with human annotators. Using the dataset labeled with weak supervision in combination with generative modeling, an F-1 score of 0.61 is achieved on the task of classifying the image contents of Instagram posts based solely on the associated text, which is on level with human performance.

  • 25.
    Hammar, Kim
    et al.
    KTH.
    Jaradat, Shatha
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Dokoohaki, Nima
    KTH.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Deep Text Mining of Instagram Data Without Strong Supervision2018In: Proceedings - 2018 IEEE/WIC/ACM International Conference on Web Intelligence, WI 2018, IEEE, 2018, p. 158-165Conference paper (Refereed)
    Abstract [en]

    With the advent of social media, our online feeds increasingly consist of short, informal, and unstructured text. This textual data can be analyzed for the purpose of improving user recommendations and detecting trends. Instagram is one of the largest social media platforms, containing both text and images. However, most of the prior research on text processing in social media is focused on analyzing Twitter data, and little attention has been paid to text mining of Instagram data. Moreover, many text mining methods rely on annotated training data, which in practice is both difficult and expensive to obtain. In this paper, we present methods for unsupervised mining of fashion attributes from Instagram text, which can enable a new kind of user recommendation in the fashion domain. In this context, we analyze a corpora of Instagram posts from the fashion domain, introduce a system for extracting fashion attributes from Instagram, and train a deep clothing classifier with weak supervision to classify Instagram posts based on the associated text. With our experiments, we confirm that word embeddings are a useful asset for information extraction. Experimental results show that information extraction using word embeddings outperforms a baseline that uses Levenshtein distance. The results also show the benefit of combining weak supervision signals using generative models instead of majority voting. Using weak supervision and generative modeling, an F1 score of 0.61 is achieved on the task of classifying the image contents of Instagram posts based solely on the associated text, which is on level with human performance. Finally, our empirical study provides one of the few available studies on Instagram text and shows that the text is noisy, that the text distribution exhibits the long-tail phenomenon, and that comment sections on Instagram are multi-lingual.

  • 26.
    Haseeb, Abdul
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Kungas, Peep
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Semantic Middleware for Robot Swarm Interaction through Web Services2008In: MIC Special Session on Computing Systems in Dynamic Environments / [ed] L. Bruzzone, 2008Conference paper (Refereed)
    Abstract [en]

    In this paper we propose a semantic middleware architecture and communication support for environments with swarm robots. Our main assumption is that robots can communicate via wireless networks while we don't assume high processing power in the robots. Basic advantage of the proposed middleware is the extension of robots' capabilities via access to semantic information and powerful processing engines. The architecture is conformant with main standard solutions and allows reusing intelligent functionality implemented in the external world.

  • 27.
    Haseeb, Abdul
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Küngas, Peep
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Mediator-based Distributed Web Services Discovery and Invocation for Infrastructure-less Mobile Dynamic Systems2008In: proceedings of 4th International IEEE Conference of Next generation Web services practices (NWeSP.08), 2008, p. 46-53Conference paper (Refereed)
    Abstract [en]

    Mobile autonomous systems like robot swarms or mobile software agents operate in a dynamic environment pertaining self-organization, self configurationand heterogeneity of computing entities.In such settings there is a need for autonomicpublishing and discovery of resources and just-in-timeintegration for on-the-fly service consumption withoutany a priori knowledge of available services both withinthe execution environment and from the outsideworld. We propose a mediator-based distributed Webservices discovery and invocation middleware.Moreover we present experimental results on animplemented robot swarm simulation environment. Wepropose a conceptual classification of computingentities on the basis of communication capabilities andconceptual overlay formation for query propagation.Our approach provides a loose coupling in terms ofspace and time and uses both Internet-basedcommunication and RDF-based communication viamessages mediators/post-boxes between entities wheninter-communication between entities is not possible.

  • 28.
    Haseeb, Abdul
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Kungas, Peep
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    DeLP Based Semantic Location Lattice for Intelligent Robotic Navigation2008In: Proceedings of the 2008 International Conference on Artificial Intelligence, ICAI 2008 and Proceedings of the 2008 International Conference on Machine Learning; Models, Technologies and Applications / [ed] H. Arabnia et al, 2008, p. 686-692Conference paper (Refereed)
    Abstract [en]

    Location models require a well-defined representation of spatial connectivity and hierarchical relationship between different spatial concepts; and are fundamental for location navigation, location based services and contextual query responses. Current location models rely on a priori knowledge of surrounding environment and mostly the semantics of relationships are over-looked. In this paper we propose an incremental semantic spatial relationship building approach for robotic agents based on formal concept analysis and defeasible reasoning. We consider a number of cases in which an autonomous robot with incomplete information about the environment can perform reasoning and update its location navigation. We use contextual information for establishing strength of partial order relationship between discovered concepts of robotic navigation/computing environment.

  • 29.
    Haseeb, Abdul
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Kungas, Peep
    Distributed and Passive Web services discovery middleware for Pervasive services at the edges of Internet2010In: Second International Conferences on Advanced Service Computing SERVICE COMPUTATION, 2010, p. 160-165Conference paper (Refereed)
  • 30.
    Haseeb, Abdul
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Kungas, Peep
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Light-Weight Decentralized Autonomic Web Service Discovery for Systems with Heterogeneous Communication Capabilities2008In: the proceedings of 12th IASTED International Conference on Internet and Multimedia Systems and Applications (IMSA.08) / [ed] M. Mandal, 2008, p. 7-18Conference paper (Refereed)
    Abstract [en]

    Interoperability between autonomous systems like robot swarm or mobile software agents rely on efficient and seamless communication. Such mobile and dynamic environments pertain self-organization and self configuration of computing entities, a need for autonomic publishing and discovery of resources, and communication from and to outside world. Furthermore, such systems are attributed by heterogeneous communication capabilities of various computing entities. We take Web services approach for robot swarm based on robotic communication capabilities and propose a collaborative and decentralized services discovery and management middleware. Our approach provides a loose coupling in terms of space and time and uses both Internet based communication and RFID tags as message post boxes/relays for communication between robots when communication over the Internet is not available.

  • 31.
    Haseeb, Abdul
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Küngas, Peep
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Distributed Discovery and Invocation of Web Services in Infrastructure-less Dynamic Environments2009In: International Journal of Web Services Practices (IJWSP), ISSN 1738-6535, Journal of Web Services Practices, ISSN 1738-6535, Vol. 3, no 3-4, p. 171-184Article in journal (Refereed)
    Abstract [en]

    Mobile autonomous systems like robot swarms or mobile software agents operate in a dynamic environment pertaining self-organization, selfconfiguration and heterogeneity of computing entities. In such settings there is a need for autonomic publishing and discovery of resources and just-in-time integration for on-the-fly service consumption without any a priori knowledge of available services both within the execution environment and from the outside world. We propose a mediator-based distributed Web services discovery and invocation middleware. Moreover we present experimental results on an implemented robot swarm simulation environment. We propose a conceptual classification of computing entities on the basis of communication capabilities and conceptual overlay formation for query propagation. Our approach provides a loose coupling in terms of space and time and uses both Internet-based communication and RDF-based communication via messages mediators/post-boxes between entities when inter-communication between entities is not possible.

  • 32.
    Haseeb, Abdul
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Küngas, Peep
    Distributed Web Services Discovery Middleware for Edges of Internet2010In: IEEE International Conference on Web Services, 2010, p. 680-682Conference paper (Refereed)
    Abstract [en]

    The advent of mobile computing devices and development of wireless and ad-hoc networking technologies has led to growth of infrastructure-less environments. Mostly, these environments lie at the edges of Internet i.e. they are disconnected/sparsely connected to rest of the world. In order to exploit the access to such edges of Internet, we propose and experimentally evaluate an interoperability middleware that synergizes P2P technology, message queuing support and a passive distributed UDDI for Web services discovery and invocation.

  • 33. Heiberg, T.
    et al.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Pedersen, J.
    An agent-based architecture for customer services management and product search2002In: Informatica (Vilnius), ISSN 0868-4952, E-ISSN 1822-8844, Vol. 13, no 4, p. 441-454Article in journal (Refereed)
    Abstract [en]

    The amount of products and services available over the Internet increases significantly and it soon becomes beyond users ability to analyze and compare them. At the same time the number of potential customers available via the Internet also increases dramatically and starts to be beyond the service providers ability to perform efficient targeted marketing. A possible way for relaxing the above-mentioned limitations could be in usage of electronic assistants, both for customers and providers. Such assistants may serve as mediators for commercial Internet-based activity. Software agents could play role of such mediators representing customers and providers in the network. In this paper we present our experience and a solution to using agent technology in customer services management for mobile users. The solution is intended to increase granularity and personalization in targeted advertising while ensuring customer privacy. The proposed solution has been implemented in a prototype system for providing services for,users of mobile devices.

  • 34.
    Jaradat, Shatha
    et al.
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Dokoohaki, Nima
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Hammar, Kim
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Wara, Ummal
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Dynamic CNN Models For Fashion Recommendation in Instagram2018In: 2018 IEEE INT CONF ON PARALLEL & DISTRIBUTED PROCESSING WITH APPLICATIONS, UBIQUITOUS COMPUTING & COMMUNICATIONS, BIG DATA & CLOUD COMPUTING, SOCIAL COMPUTING & NETWORKING, SUSTAINABLE COMPUTING & COMMUNICATIONS / [ed] Chen, JJ Yang, LT, IEEE COMPUTER SOC , 2018, p. 1144-1151Conference paper (Refereed)
    Abstract [en]

    Instagram as an online photo-sharing and social-networking service is becoming more powerful in enabling fashion brands to ramp up their business growth. Nowadays, a single post by a fashion influencer attracts a wealth of attention and a magnitude of followers who are curious to know more about the brands and style of each clothing item sitting inside the image. To this end, the development of efficient Deep CNN models that can accurately detect styles and brands have become a research challenge. In addition, current techniques need to cope with inherent fashion-related data issues. Namely, clothing details inside a single image only cover a small proportion of the large and hierarchical space of possible brands and clothing item attributes. In order to cope with these challenges, one can argue that neural classifiers should become adapted to large-scale and hierarchical fashion datasets. As a remedy, we propose two novel techniques to incorporate the valuable social media textual content to support the visual classification in a dynamic way. The first method is adaptive neural pruning (DynamicPruning) in which the clothing item category detected from posts' text analysis can be used to activate the possible range of connections of clothing attributes' classifier. The second method (DynamicLayers) is a dynamic framework in which multiple-attributes classification layers exist and a suitable attributes' classifier layer is activated dynamically based upon the mined text from the image. Extensive experiments on a dataset gathered from Instagram and a baseline fashion dataset (DeepFashion) have demonstrated that our approaches can improve the accuracy by about 20% when compared to base architectures. It is worth highlighting that with Dynamiclayers we have gained 35% accuracy for the task of multi-class multi-labeled classification compared to the other model.

  • 35.
    Jaradat, Shatha
    et al.
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Dokoohaki, Nima
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    OLLDA: A Supervised and Dynamic Topic Mining Framework in Twitter2015In: 2015 IEEE International Conference on Data Mining Workshop (ICDMW), 2015, p. 1354-1359Conference paper (Refereed)
    Abstract [en]

    Analyzing media in real-time is of great importance with social media platforms at the epicenter of crunching, digesting and disseminating content to individuals connected to these platforms. Within this context, topic models, specially LDA, have gained strong momentum due to their scalability, inference power and their compact semantics. Although, state of the art topic models come short in handling streaming large chunks of data arriving dynamically onto the platform, thus hindering their quality of interpretation as well as their adaptability to information overload. As a result, in this manuscript we propose for a labelled and online extension to LDA (OLLDA), which incorporates supervision through external labeling and capability of quickly digesting real-time updates thus making it more adaptive to Twitter and platforms alike. Our proposed extension has capability of handling large quantities of newly arrived documents in a stream, and at the same time, is capable of achieving high topic inference quality given the short and often sloppy text of tweets. Our approach mainly uses an approximate inference technique based on variational inference coupled with a labeled LDA model. We conclude by presenting experiments using a one year crawl of Twitter data that shows significantly improved topical inference as well as temporal user profile classification when compared to state of the art baselines.

  • 36.
    Jaradat, Shatha
    et al.
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Dokoohaki, Nima
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Outfit2Vec: Incorporating Clothing Hierarchical MetaData into Outfits’ Recommendation2020In: Fashion Recommender Systems, Springer International Publishing , 2020, p. 87-107Chapter in book (Refereed)
    Abstract [en]

    Fashion Personalisation is emerging as a major service that online retailers and brands are competing to provide. They aim to deliver more tailored recommendations to increase revenues and satisfy customers by providing them options of similar items according to their purchase history. However, many online retailers still struggle with turning customers’ data into actionable and intelligent recommendations that reflect their personalised and preferred taste of style. On the other hand due to the ever increasing use of social media, fashion brands invest in influencers’ marketing to advertise their brands to reach a larger segment of customers who strongly trust their influencers’ choices. In this context the textual and visual analysis of social media can be used to extract semantic knowledge about customers’ preferences that can be further applied in generating tailored online shopping recommendations. As style lies in the details of outfits, recommendation models should leverage the fashion metadata ranging from clothing categories and subcategories to attributes such as materials and patterns to overall style description in order to generate fine-grained recommendations. Recently, several recommendation algorithms suggested to model the latent representations of items and users with neural word embeddings approaches which showed improved results. Inspired by Paragraph Vector neural embeddings model, we present Outfit2vec and PartialOutfit2vec in which we leverage the complex relationship between user’s fashion metadata while generating outfits’ embeddings. In this paper, we also describe a methodology to generate representative vectors of hierarchically-composed fashion outfits. We evaluate our models using different strategies in comparison to the paragraph embedding models on an extensively-annotated Instagram dataset on recommendation and multi-class style classification tasks. Our models achieve better results specially in whole outfits’ ranking evaluations with an average of 22% increase.

  • 37.
    Jaradat, Shatha
    et al.
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Dokoohaki, Nima
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Outfit2Vec: Incorporating Clothing Hierarchical MetaData into Outfits’ Recommendation2020In: Lecture Notes in Social Networks book series (LNSN): Fashion Recommender Systems / [ed] Springer, Springer: Springer Nature , 2020Chapter in book (Other academic)
  • 38.
    Jaradat, Shatha
    et al.
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Dokoohaki, Nima
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Ferrari, Elena
    Learning What to Share in Online Social Networks Using Deep Reinforcement Learning2018In: Machine Learning Techniques for Online Social Networks, Springer International Publishing , 2018, p. 115-133Chapter in book (Other academic)
    Abstract [en]

    Online networking sites tried their best to have right privacy mechanisms in place for users, enabling them to share the right content with the right audience. With all these efforts, privacy customizations remain hard for users across the sites. Existing research that addresses this problem mainly focuses on semi-supervised strategies that introduce extra complexity by requiring the user to manually specify initial privacy preferences for their friends. In this work, we suggest a deep reinforcement learning framework that can dynamically generate privacy labels for users in OSNs. We evaluated our framework on a 1 year crawl of Twitter data, using different types of recurrent units in recurrent neural networks (RNN): Long Short Term Memory (LSTM), Gated Recurrent Unit (GRU), and Simple RNN. Our experiments revealed that LSTM performed better than GRU in terms of top users detection accuracy and the ranked dependence between the generated privacy labels and estimated user trust values.

  • 39.
    Jaradat, Shatha
    et al.
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Dokoohaki, Nima
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Ferrari, Elena
    Trust And Privacy Correlations in Social Networks: A Deep Learning Framework2016In: PROCEEDINGS OF THE 2016 IEEE/ACM INTERNATIONAL CONFERENCE ON ADVANCES IN SOCIAL NETWORKS ANALYSIS AND MINING ASONAM 2016, IEEE conference proceedings, 2016, p. 203-206Conference paper (Refereed)
    Abstract [en]

    Online Social Networks (OSNs) remain the focal point of Internet usage. Since the beginning, networking sites tried best to have right privacy mechanisms in place for users, enabling them to share the right content with the right audience. With all these efforts, privacy customizations remain hard for users across the sites. Existing research that address this problem mainly focus on semi-supervised strategies that introduce extra complexity by requiring the user to manually specify initial privacy preferences for their friends. In this work, we suggest an adaptive solution that can dynamically generate privacy labels for users in OSNs. To this end, we introduce a deep reinforcement learning framework that targets two key problems in OSNs like Facebook: the exposure of users' interactions through the network to less trusted direct friends, and the possibility of propagating user updates through direct friends' interactions to indirect friends. By implementing this framework, we aim at understanding how social trust and privacy could be correlated, specifically in a dynamic fashion. We report the ranked dependence between the generated privacy labels and the estimated user trust values, which indicate the ability of the framework to identify the highly trusted users and share with them higher percentages of data.

  • 40.
    Jaradat, Shatha
    et al.
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Dokoohaki, Nima
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Wara, Ummal
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Goswami, Mallu
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Hammar, Kim
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    TALS: A framework for text analysis, fine-grained annotation, localisation and semantic segmentation2019In: Proceedings - International Computer Software and Applications Conference, IEEE Computer Society, 2019, Vol. 8754470, p. 201-206Conference paper (Refereed)
    Abstract [en]

    With around 2.77 billion users using online social media platforms nowadays, it is becoming more attractive for business retailers to reach and to connect to more potential clients through social media. However, providing more effective recommendations to grab clients’ attention requires a deep understanding of users’ interests. Given the enormous amounts of text and images that users share in social media, deep learning approaches play a major role in performing semantic analysis of text and images. Moreover, object localisation and pixel-by-pixel semantic segmentation image analysis neural architectures provide an enhanced level of information. However, to train such architectures in an end-to-end manner, detailed datasets with specific meta-data are required. In our paper, we present a complete framework that can be used to tag images in a hierarchical fashion, and to perform object localisation and semantic segmentation. In addition to this, we show the value of using neural word embeddings in providing additional semantic details to annotators to guide them in annotating images in the system. Our framework is designed to be a fully functional solution capable of providing fine-grained annotations, essential localisation and segmentation services while keeping the core architecture simple and extensible. We also provide a fine-grained labelled fashion dataset that can be a rich source for research purposes.

  • 41.
    Javed, Mohtasim
    et al.
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Haseeb, Abdul
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Semantic manipulation of non- Semantic/natural-language queries2010In: Proceedings of the 9th Joint Conference on Knowledge-Based Software Engineering, JCKBSE 2010, 2010, p. 210-222Conference paper (Refereed)
    Abstract [en]

    Rapid growth of information has made information access inherently difficult. Semantic Web, with the help of formal Ontologies, has provided an efficient way to represent data which makes it easy to find and organize by creating a mesh of information linked up in machine process-able way. The formal query languages (SPARQL) can exploit the semantics in information access; however formal languages are complex for naïve users. On the other hand, keyword search can't benefit from semantics of information. In this paper we present a query interface that performs semantic manipulation of non- semantic natural language-based queries. The proposed query interface semantically manipulates the user query to generate RDF triples to be queried using SPARQL. Proposed system provides the richness of semantics of formal query languages while maintaining query simplicity.

  • 42. Khan, A. B.
    et al.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Software and Computer systems, SCS.
    Supporting place/space based patterns of citywide mobile learning through multi-agent framework2012In: Proceedings 2012 17th IEEE International Conference on Wireless, Mobile and Ubiquitous Technology in Education, WMUTE 2012, IEEE , 2012, p. 152-156Conference paper (Refereed)
    Abstract [en]

    There are numerous possible patterns consisting of learners, technological components and physical locations (i.e. Spaces) which can be identified in the context of citywide mobile learning (CML). By patterns, we mean to refer to the different ways in which learners are associated with the Spaces where the learning occurs, with technology that supports learning and among the learners themselves. Envisioning the entire set of possible scenarios, which can exist in the citywide context and designing to support them is not only difficult, but also practically impossible. Therefore, there is a need to condense and generalize all the numerous possible scenarios of CML into few core patterns, which can be use as basic building blocks to construct and support more complex CML scenarios. A regular trend in the currently existing literature is to consider only one or two use-cases while neglecting others. In this paper we address this problem by presenting six general patterns instead of use-cases that can exist in citywide context. These patterns take into account three fundamental aspects of CML, (1) learner, (2) Place/Space and (3) the technological components needed to support CML.

  • 43.
    Khan, Akif Quddus
    et al.
    Norwegian University of Science and Technology – NTNU, Gjøvik, Norway.
    Nikolov, Nikolay
    SINTEF AS, Oslo, Norway.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Prodan, Radu
    University of Klagenfurt, Klagefurt, Austria.
    Bussler, Christoph
    Robert Bosch LLC, Sunnyvale, CA, USA.
    Roman, Dumitru
    SINTEF AS, Oslo, Norway; OsloMet – Oslo Metropolitan University, Oslo, Norway.
    Soylu, Ahmet
    OsloMet – Oslo Metropolitan University, Oslo, Norway.
    Towards Cloud Storage Tier Optimization with Rule-Based Classification2023In: Service-Oriented and Cloud Computing: 10th IFIP WG 6.12 European Conference, ESOCC 2023, Proceedings, Springer Nature , 2023, p. 205-216Conference paper (Refereed)
    Abstract [en]

    Cloud storage adoption has increased over the years as more and more data has been produced with particularly high demand for fast processing and low latency. To meet the users’ demands and to provide a cost-effective solution, cloud service providers (CSPs) have offered tiered storage; however, keeping the data in one tier is not a cost-effective approach. Hence, several two-tiered approaches have been developed to classify storage objects into the most suitable tier. In this respect, this paper explores a rule-based classification approach to optimize cloud storage cost by migrating data between different storage tiers. Instead of two, four distinct storage tiers are considered, including premium, hot, cold, and archive. The viability and potential of the approach are demonstrated by comparing cost savings achieved when data was moved between tiers versus when it remained static. The results indicate that the proposed approach has the potential to significantly reduce cloud storage cost, thereby providing valuable insights for organizations seeking to optimize their cloud storage strategies. Finally, the limitations of the proposed approach are discussed along with the potential directions for future work, particularly the use of game theory to incorporate a feedback loop to extend and improve the proposed approach accordingly.

  • 44.
    Khan, Akif Quddus
    et al.
    Norwegian University of Science and Technology, Norway.
    Nikolov, Nikolay
    SINTEF Digital, Norway.
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Prodan, Radu
    University of Klagenfurt, Austria.
    Bussler, Christoph
    Robert Bosch LLC, CA, USA.
    Roman, Dumitru
    SINTEF Digital, Norway.
    Soylu, Ahmet
    OsloMet - Oslo Metropolitan University, Norway.
    Towards Graph-based Cloud Cost Modelling and Optimisation2023In: Proceedings: 2023 IEEE 47th Annual Computers, Software, and Applications Conference, COMPSAC 2023, Institute of Electrical and Electronics Engineers (IEEE) , 2023, p. 1337-1342Conference paper (Refereed)
    Abstract [en]

    Cloud computing has become an increasingly popular choice for businesses and individuals due to its flexibility, scalability, and convenience; however, the rising cost of cloud resources has become a significant concern for many. The pay-per-use model used in cloud computing means that costs can accumulate quickly, and the lack of visibility and control can result in unexpected expenses. The cost structure becomes even more complicated when dealing with hybrid or multi-cloud environments. For businesses, the cost of cloud computing can be a significant portion of their IT budget, and any savings can lead to better financial stability and competitiveness. In this respect, it is essential to manage cloud costs effectively. This requires a deep understanding of current resource utilization, forecasting future needs, and optimising resource utilization to control costs. To address this challenge, new tools and techniques are being developed to provide more visibility and control over cloud computing costs. In this respect, this paper explores a graph-based solution for modelling cost elements and cloud resources and potential ways to solve the resulting constraint problem of cost optimisation. We primarily consider utilization, cost, performance, and availability in this context. Such an approach will eventually help organizations make informed decisions about cloud resource placement and manage the costs of software applications and data workflows deployed in single, hybrid, or multi-cloud environments.

  • 45.
    Khan, Akif Quddus
    et al.
    Norwegian Univ Sci & Technol NTNU, Dept Comp Sci, N-2815 Gjovik, Norway..
    Nikolov, Nikolay
    SINTEF AS, SINTEF Digital, N-0373 Oslo, Norway..
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Prodan, Radu
    Univ Klagenfurt, Dept Informat Technol, A-9020 Klagenfurt, Austria..
    Roman, Dumitru
    SINTEF AS, SINTEF Digital, N-0373 Oslo, Norway..
    Sahin, Bekir
    Natl Univ Sci & Technol, Logist Management, Sohar 111, Oman..
    Bussler, Christoph
    Robert Bosch LLC, Sunnyvale, CA 94085 USA..
    Soylu, Ahmet
    OsloMet Oslo Metropolitan Univ, Dept Comp Sci, N-0167 Oslo, Norway..
    Smart Data Placement Using Storage-as-a-Service Model for Big Data Pipelines2023In: Sensors, E-ISSN 1424-8220, Vol. 23, no 2, article id 564Article in journal (Refereed)
    Abstract [en]

    Big data pipelines are developed to process data characterized by one or more of the three big data features, commonly known as the three Vs (volume, velocity, and variety), through a series of steps (e.g., extract, transform, and move), making the ground work for the use of advanced analytics and ML/AI techniques. Computing continuum (i.e., cloud/fog/edge) allows access to virtually infinite amount of resources, where data pipelines could be executed at scale; however, the implementation of data pipelines on the continuum is a complex task that needs to take computing resources, data transmission channels, triggers, data transfer methods, integration of message queues, etc., into account. The task becomes even more challenging when data storage is considered as part of the data pipelines. Local storage is expensive, hard to maintain, and comes with several challenges (e.g., data availability, data security, and backup). The use of cloud storage, i.e., storage-as-a-service (StaaS), instead of local storage has the potential of providing more flexibility in terms of scalability, fault tolerance, and availability. In this article, we propose a generic approach to integrate StaaS with data pipelines, i.e., computation on an on-premise server or on a specific cloud, but integration with StaaS, and develop a ranking method for available storage options based on five key parameters: cost, proximity, network performance, server-side encryption, and user weights/preferences. The evaluation carried out demonstrates the effectiveness of the proposed approach in terms of data transfer performance, utility of the individual parameters, and feasibility of dynamic selection of a storage option based on four primary user scenarios.

  • 46.
    Khan, Akif Quddus
    et al.
    Norwegian Univ Sci & Technol, Trondheim, Norway..
    Nikolov, Nikolay
    SINTEF Digital, Oslo, Norway..
    Matskin, Mihhail
    KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Software and Computer systems, SCS.
    Prodan, Radu
    Univ Klagenfurt, Klagenfurt, Austria..
    Song, Hui
    SINTEF Digital, Oslo, Norway..
    Roman, Dumitru
    SINTEF Digital, Oslo, Norway..
    Soylu, Ahmet
    OsloMet Oslo Metropolitan Univ, Oslo, Norway..
    Smart Data Placement for Big Data Pipelines: An Approach based on the Storage-as-a-Service Model2022In: 2022 IEEE/ACM 15TH INTERNATIONAL CONFERENCE ON UTILITY AND CLOUD COMPUTING, UCC, Institute of Electrical and Electronics Engineers (IEEE) , 2022, p. 317-320Conference paper (Refereed)
    Abstract [en]

    The development of big data pipelines is a challenging task, especially when data storage is considered as part of the data pipelines. Local storage is expensive, hard to maintain, comes with several challenges (e.g., data availability, data security, and backup). The use of cloud storage, i.e., Storageas-a-Service (StaaS), instead of local storage has the potential of providing more flexibility in terms of such as scalability, fault tolerance, and availability. In this paper, we propose a generic approach to integrate StaaS with data pipelines, i.e., computation on an on-premise server or on a specific cloud, but integration with StaaS, and develop a ranking method for available storage options based on five key parameters: cost, proximity, network performance, the impact of server-side encryption, and user weights. The evaluation carried out demonstrates the effectiveness of the proposed approach in terms of data transfer performance and the feasibility of dynamic selection of a storage option based on four primary user scenarios.

  • 47. Khan, Basit
    et al.
    Matskin, Mihhail
    Norwegian University of Science and Technology, Norway.
    A Platform for Actively Supporting e-Learning in Mobile Networks2010In: International Journal of Mobile and Blended Learning, ISSN 1941-8647, E-ISSN 1941-8655, International Journal of Mobile and Blended Learning, ISSN 1941-8647, Vol. 2, p. 55-79Article in journal (Refereed)
  • 48.
    Khan, Basit
    et al.
    Department of Computer and Information Science (IDI) Norwegian University of Science and Technology (NTNU) Trondheim, Norway.
    Matskin, Mihhail
    Department of Computer and Information Science (IDI) Norwegian University of Science and Technology (NTNU) Trondheim, Norway.
    AGORA Framework for Service Discovery and Resource Allocation2010In: The Fifth International Conference on Internet and Web Applications and Services ICIW 2010, 2010, p. 438-444Conference paper (Refereed)
  • 49. Khan, Basit
    et al.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    FABULA Platform for Active e-Learning in Mobile Networks2009In: Proceedings of the IADIS International Conference Mobile Learning 2009., 2009, p. 33-41Conference paper (Refereed)
  • 50. Khan, Basit
    et al.
    Matskin, Mihhail
    KTH, School of Information and Communication Technology (ICT), Communication: Services and Infrastucture, Software and Computer Systems, SCS.
    Multiagent system to support Place/Space based mobile learning in city2011In: International Conference on Information Society, i-Society 2011, 2011, p. 66-71Conference paper (Refereed)
    Abstract [en]

    Different approaches have been developed to provide technical support for mobile learning. Most these approaches consider only the physical properties of learning environment. In this work, we not only focus on the physical/spatial dimension of the learning environment of the city, but also pay attention to the notion of Place which is a meaningful outcome of peoples understanding of Space. This paper illustrates how a theoretical conceptualization of Spaces and Places is mapped into a multiagent framework called AGORA. It presents the design aspects of a mobile learning system, which uses software agents as its core functional units. We discuss how the theoretical concepts are used to define a technical solution to support mobile learning in a citywide context.

123 1 - 50 of 132
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf