Ändra sökning
Avgränsa sökresultatet
1 - 48 av 48
RefereraExporteraLänk till träfflistan
Permanent länk
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
Träffar per sida
  • 5
  • 10
  • 20
  • 50
  • 100
  • 250
Sortering
  • Standard (Relevans)
  • Författare A-Ö
  • Författare Ö-A
  • Titel A-Ö
  • Titel Ö-A
  • Publikationstyp A-Ö
  • Publikationstyp Ö-A
  • Äldst först
  • Nyast först
  • Skapad (Äldst först)
  • Skapad (Nyast först)
  • Senast uppdaterad (Äldst först)
  • Senast uppdaterad (Nyast först)
  • Disputationsdatum (tidigaste först)
  • Disputationsdatum (senaste först)
  • Standard (Relevans)
  • Författare A-Ö
  • Författare Ö-A
  • Titel A-Ö
  • Titel Ö-A
  • Publikationstyp A-Ö
  • Publikationstyp Ö-A
  • Äldst först
  • Nyast först
  • Skapad (Äldst först)
  • Skapad (Nyast först)
  • Senast uppdaterad (Äldst först)
  • Senast uppdaterad (Nyast först)
  • Disputationsdatum (tidigaste först)
  • Disputationsdatum (senaste först)
Markera
Maxantalet träffar du kan exportera från sökgränssnittet är 250. Vid större uttag använd dig av utsökningar.
  • 1.
    Abbas, Zainab
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Al-Shishtawy, Ahmad
    RISE SICS, Stockholm, Sweden.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS. RISE SICS, Stockholm, Sweden..
    Vlassov, Vladimir
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Short-Term Traffic Prediction Using Long Short-Term Memory Neural Networks2018Konferensbidrag (Refereegranskat)
    Abstract [en]

    Short-term traffic prediction allows Intelligent Transport Systems to proactively respond to events before they happen. With the rapid increase in the amount, quality, and detail of traffic data, new techniques are required that can exploit the information in the data in order to provide better results while being able to scale and cope with increasing amounts of data and growing cities. We propose and compare three models for short-term road traffic density prediction based on Long Short-Term Memory (LSTM) neural networks. We have trained the models using real traffic data collected by Motorway Control System in Stockholm that monitors highways and collects flow and speed data per lane every minute from radar sensors. In order to deal with the challenge of scale and to improve prediction accuracy, we propose to partition the road network into road stretches and junctions, and to model each of the partitions with one or more LSTM neural networks. Our evaluation results show that partitioning of roads improves the prediction accuracy by reducing the root mean square error by the factor of 5. We show that we can reduce the complexity of LSTM network by limiting the number of input sensors, on average to 35% of the original number, without compromising the prediction accuracy.

  • 2. Aberer, Karl
    et al.
    Alima, Luc Onana
    Ghodsi, Ali
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Elektronik- och datorsystem, ECS.
    Girdzijauskas, Sarunas
    Ecole Polytechnique Fédérale de Lausanne.
    Haridi, Seif
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Elektronik- och datorsystem, ECS.
    Hauswirth, Manfred
    The Essence of P2P: A Reference Architecture for Overlay Networks2005Ingår i: Fifth IEEE International Conference on Peer-to-Peer Computing, Proceedings / [ed] Caronni, G; Weiler, N; Waldvogel, M; Shahmehri, N, 2005, s. 11-20Konferensbidrag (Refereegranskat)
    Abstract [en]

    The success of the P2P idea has created a huge diversity of approaches, among which overlay networks, for example, Gnutella, Kazaa, Chord, Pastry, Tapestry, P-Grid, or DKS, have received specific attention from both developers and researchers. A wide variety of algorithms, data structures, and architectures have been proposed. The terminologies and abstractions used, however have become quite inconsistent since the P2P paradigm has attracted people from many different communities, e.g., networking, databases, distributed systems, graph theory, complexity theory, biology, etc. In this paper we propose a reference model for overlay networks which is capable of modeling different approaches in this domain in a generic manner It is intended to allow researchers and users to assess the properties of concrete systems, to establish a common vocabulary for scientific discussion, to facilitate the qualitative comparison of the systems, and to serve as the basis for defining a standardized API to make overlay networks interoperable.

  • 3.
    Apolonia, Nuno
    et al.
    Universitat Politecnica de Catalunya (UPC) Barcelona, Spain.
    Antaris, Stefanos
    Girdzijauskas, Šarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Pallis, G.
    Dikaiakos, Marios
    SELECT: A distributed publish/subscribe notification system for online social networks2018Ingår i: Proceedings - 2018 IEEE 32nd International Parallel and Distributed Processing Symposium, IPDPS 2018, Institute of Electrical and Electronics Engineers (IEEE), 2018, s. 970-979, artikel-id 8425250Konferensbidrag (Refereegranskat)
    Abstract [en]

    Publish/subscribe (pub/sub) mechanisms constitutean attractive communication paradigm in the design of large-scale notification systems for Online Social Networks (OSNs). Toaccommodate the large-scale workloads of notifications producedby OSNs, pub/sub mechanisms require thousands of serversdistributed on different data centers all over the world, incurringlarge overheads. To eliminate the pub/sub resources used, wepropose SELECT - a distributed pub/sub social notificationsystem over peer-to-peer (P2P) networks. SELECT organizesthe peers on a ring topology and provides an adaptive P2Pconnection establishment algorithm where each peer identifiesthe number of connections required, based on the social structureand user availability. This allows to propagate messages to thesocial friends of the users using a reduced number of hops.The presented algorithm is an efficient heuristic to an NP-hard problem which maps workload graphs to structured P2Poverlays inducing overall, close to theoretical, minimal number ofmessages. Experiments show that SELECT reduces the numberof relay nodes up to 89% versus the state-of-the-art pub/subnotification systems. Additionally, we demonstrate the advantageof SELECT against socially-aware P2P overlay networks andshow that the communication between two socially connectedpeers is reduced on average by at least 64% hops, while achieving100% communication availability even under high churn.

  • 4.
    Apolonia, Nuno
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT). Universitat Politecnica de Catalunya (UPC) Barcelona, Spain.
    Freitag, Felix
    Universitat Politècnica de Catalunya. Barcelona, Spain.
    Navarro, Leandro
    Universitat Politècnica de Catalunya, BarcelonaTECH, Barcelona, Spain.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Socially aware microcloud service overlay optimization in community networks2019Ingår i: Software, practice & experience, ISSN 0038-0644, Vol. 49, nr 1, artikel-id 13Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    Community networks are a growing network cooperation effort by citizens to build and maintain Internet infrastructure in regions that are not available. Adding that, to bring cloud services to community networks (CNs), microclouds were started as an edge cloud computing model where members cooperate using resources. Therefore, enhancing routing for services in CNs is an attractive paradigm that benefits the infrastructure. The problem is the growing consumption of resources for disseminating messages in the CN environment. This is because the services that build their overlay networks are oblivious to the underlying workload patterns that arise from social cooperation in CNs. In this paper, we propose Select in Community Networks (SELECTinCN), which enhances the overlay creation for pub/sub systems over peer‐to‐peer (P2P) networks. Moreover, SELECTinCN includes social information based on cooperation within CNs by exploiting the social aspects of the community of practice. Our work organizes the peers in a ring topology and provides an adaptive P2P connection establishment algorithm, where each peer identifies the number of connections needed based on the social structure and user availability. This allows us to propagate messages using a reduced number of hops, thus providing an efficient heuristic to an NP‐hard problem that maps the workload graph to the structured P2P overlays resulting in a number of messages close to the theoretical minimum. Experiments show that, by using social network information, SELECTinCN reduces the number of relay nodes by up to 89% using the community of practice information versus the state‐of‐the‐art pub/sub notification systems given as baseline.

  • 5.
    Bahri, Leila
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Blockchain technology: Practical P2P computing (Tutorial)2019Ingår i: Proceedings - 2019 IEEE 4th International Workshops on Foundations and Applications of Self* Systems, FAS*W 2019, Institute of Electrical and Electronics Engineers (IEEE), 2019, s. 249-250, artikel-id 8791982Konferensbidrag (Refereegranskat)
    Abstract [en]

    Blockchain technology comes with the promise to revolutionize the way current IT systems are organized as well as to revise how trust is perceived in the wider society. In spite of the wide attention that cyrpto-currencies (such as Bitcoin) have attracted, Blockchain technology is more likely to make an impact beyond ongoing speculations on cyrpto-currencies. Decentralized identity management, transparent supply-chain systems, and IoT governance and security are only few examples of research challenges for which this technology may hold substantial potential. Blockchain technology has emerged at the intersection of two well established research areas: peer-to-peer (P2P) computing and cryptography. In this tutorial, we provide a general overview of the main components behind this technology, we present the difference between the types of Blockchain available today, and we make a high level discussion on its potentials and limitations as well as possible research challenges.

  • 6.
    Bahri, Leila
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Trust Mends Blockchains: Living up to Expectations2019Ingår i: IEEE 39th International Conference on Distributed Computing Systems (ICDCS), Dallas, July 7-10 2019, 2019Konferensbidrag (Refereegranskat)
    Abstract [en]

    At the heart of Blockchains is the trustless leader election mechanism for achieving consensus among pseudoanonymous peers, without the need of oversight from any third party or authority whatsoever. So far, two main mechanisms are being discussed: proof-of-work (PoW) and proof-of-stake (PoS). PoW relies on demonstration of computational power, and comes with the markup of huge energy wastage in return of the stake in cyrpto-currency. PoS tries to address this by relying on owned stake (i.e., amount of crypto-currency) in the system. In both cases, Blockchains are limited to systems with financial basis. This forces non-crypto-currency Blockchain applications to resort to “permissioned” setting only, effectively centralizing the system. However, non-crypto-currency permisionless blockhains could enable secure and self-governed peer-to-peer structures for numerous emerging application domains, such as education and health, where some trust exists among peers. This creates a new possibility for valuing trust among peers and capitalizing it as the basis (stake) for reaching consensus. In this paper we show that there is a viable way for permisionless non-financial Blockhains to operate in completely decentralized environments and achieve leader election through proof-of-trust (PoT). In our PoT construction, peer trust is extracted from a trust network that emerges in a decentralized manner and is used as a waiver for the effort to be spent for PoW, thus dramatically reducing total energy expenditure of the system. Furthermore, our PoT construction is resilient to the risk of small cartels monopolizing the network (as it happens with the mining-pool phenomena in PoW) and is not vulnerable to sybils. We evluate security guarantees, and perform experimental evaluation of our construction, demonstrating up to 10-fold energy savings compared to PoW without trading off any of the decentralization characteristics, with further guarantees against risks of monopolization.

  • 7.
    Bahri, Leila
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    When Trust Saves Energy - A Reference Framework for Proof-of-Trust (PoT) Blockchains2018Ingår i: WWW '18 Companion Proceedings of the The Web Conference 2018, ACM Digital Library, 2018, s. 1165-1169Konferensbidrag (Refereegranskat)
    Abstract [en]

    Blockchains are attracting the attention of many technical, financial, and industrial parties, as a promising infrastructure for achieving secure peer-to-peer (P2P) transactional systems. At the heart of blockchains is proof-of-work (PoW), a trustless leader election mechanism based on demonstration of computational power. PoW provides blockchain security in trusless P2P environments, but comes at the expense of wasting huge amounts of energy. In this research work, we question this energy expenditure of PoW under blockchain use cases where some form of trust exists between the peers. We propose a Proof-of-Trust (PoT) blockchain where peer trust is valuated in the network based on a trust graph that emerges in a decentralized fashion and that is encoded in and managed by the blockchain itself. This trust is then used as a waiver for the difficulty of PoW; that is, the more trust you prove in the network, the less work you do.

  • 8. Bahri, Leila
    et al.
    Soliman, Amira
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Squillaci, Jacopo
    Carminati, Barbara
    Ferrari, Elena
    Girdzijauskas, Sarunas
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Beat the DIVa: Decentralized Identity Validation for Online Social Networks2016Ingår i: 2016 32ND IEEE INTERNATIONAL CONFERENCE ON DATA ENGINEERING (ICDE), 2016, s. 1330-1333Konferensbidrag (Refereegranskat)
    Abstract [en]

    Fake accounts in online social networks (OSNs) have known considerable sophistication and are now attempting to gain network trust by infiltrating within honest communities. Honest users have limited perspective on the truthfulness of new online identities requesting their friendship. This facilitates the task of fake accounts in deceiving honest users to befriend them. To address this, we have proposed a model that learns hidden correlations between profile attributes within OSN communities, and exploits them to assist users in estimating the trustworthiness of new profiles. To demonstrate our method, we suggest, in this demo, a game application through which players try to cheat the system and convince nodes in a simulated OSN to befriend them. The game deploys different strategies to challenge the players and to reach the objectives of the demo. These objectives are to make participants aware of how fake accounts can infiltrate within their OSN communities, to demonstrate how our suggested method could aid in mitigating this threat, and to eventually strengthen our model based on the data collected from the moves of the players.

  • 9.
    Chen, Chen
    et al.
    Middleware System Research Group, University of Toronto.
    Tock, Yoav
    IBM Research - Haifa.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    BeaConvey: Co-Design of Overlay and Routing for Topic-basedPublish/Subscribe on Small-World Networks2018Konferensbidrag (Refereegranskat)
  • 10.
    Datta, Anwitaman
    et al.
    Swiss Federal Institute of Technology Lausanne.
    Girdzijauskas, Sarunas
    Swiss Federal Institute of Technology Lausanne.
    Aberer, Karl
    Swiss Federal Institute of Technology Lausanne.
    On de Bruijn routing in distributed hash tables: There and back again2004Ingår i: The 4th IEEE International Conference on Peer-to-Peer Computing, proceedings, 2004, s. 159-166Konferensbidrag (Refereegranskat)
    Abstract [en]

    We show in this paper that de Bruijn networks, despite providing efficient search while using constant routing table size, as well as simplicity of the understanding and implementation of such networks, are unsuitable where key distribution will be uneven, a realistic scenario for most practical applications. In presence of arbitrarily skewed data distribution, it has only recently been shown that some traditional P2P overlay networks with non-constant (typically logarithmic) instead of constant routing table size can meet conflicting objectives of storage load balancing as well as search efficiency. So this paper while showing that de Bruijn networks fail, to meet these dual objectives, opens up a more general problem for the research community as to whether P2P systems with constant routing table can at all achieve the conflicting objectives of retaining search efficiency as well as storage load balancing, while preserving key ordering (which leads to uneven key distribution).

  • 11.
    Ghoorchian, Kambiz
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Spatio-Temporal Multiple Geo-Location Identification on Twitter2018Ingår i: Proceedings - 2018 IEEE International Conference on Big Data, Big Data 2018 / [ed] Abe, N Liu, H Pu, C Hu, X Ahmed, N Qiao, M Song, Y Kossmann, D Liu, B Lee, K Tang, J He, J Saltz, J, Institute of Electrical and Electronics Engineers (IEEE), 2018, s. 3412-3421Konferensbidrag (Refereegranskat)
    Abstract [en]

    Twitter Geo-tags that indicate the exact location of messages have many applications from localized opinion mining during elections to efficient traffic management in critical situations. However, less than 6% of Tweets are Geo-tagged, which limits the implementation of those applications. There are two groups of solutions: content and network-based. The first group uses location indicative factors like URLs and topics, extracted from the content of tweets, to infer Geo-location for non geoactive users, whereas the second group benefits from friendship ties in the underlying social network graph. Friendship ties are better predictors compared to content information because they are less noisy and often follow the natural human spatial movement patterns. However, their prediction's accuracy is still limited because they ignore the temporal aspects of human behavior and always assume a single location per user. This research aims to extend the current network-based approaches by taking users' temporal dimension into account. We assume multiple locations per user during different time-slots and hypothesize that location predictability varies depending on the time and the properties of the social membership group. Thus, we propose a hierarchical solution to apply temporal categorizations on top of social network partitioning for multiple location prediction for users in Online Social Networks (OSNs) like Twitter. Given a largescale Twitter dataset, we show that users' location predictability exhibits different behavior in different time-slots and different social groups. We find that there are specific conditions where users are more predictable in terms of Geo-location. Our solution outperforms the state-of-the-art by improving the prediction accuracy by 16:6% in terms of Median Error Distance (MED) over the same recall.

  • 12.
    Ghoorchian, Kambiz
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS. Swedish Institute of Technology (SICS).
    Girdzijauskas, Sarunas
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Rahimian, Fatemeh
    DeGPar: Large Scale Topic Detection usingNode-Cut Partitioning on Dense Weighted Graphs2017Ingår i: 2017 IEEE 37th International Conference on Distributed Computing Systems (ICDCS), IEEE conference proceedings, 2017, s. 775-785, artikel-id 7980020Konferensbidrag (Refereegranskat)
    Abstract [en]

    Topic Detection (TD) refers to automatic techniques for locating topically related material in web documents. Nowadays, massive amounts of documents are generated by users of Online Social Networks (OSNs), in form of very short text, tweets and snippets of news. While topic detection, in its traditional form, is applied to a few documents containing a lot of information, the problem has now changed to dealing with massive number of documents with very little information. The traditional solutions, thus, fall short either in scalability (due to huge number of input items) or sparsity (due to insufficient information per input item). In this paper we address the scalability problem by introducing an efficient and scalable graph based algorithm for TD on short texts, leveraging dimensionality reduction and clustering techniques. We first, compress the input set of documents into a dense graph, such that frequent co-occurrence patterns in the documents create multiple dense topological areas in the graph. Then, we partition the graph into multiple dense sub-graphs, each representing a topic. We compare the accuracy and scalability of our solution with two state-of-the-art solutions (including the standard LDA, and BiTerm). The results on two widely used benchmark datasets show that our algorithm not only maintains a similar or better accuracy, but also performs by an order of magnitude faster than the state-of-the-art approaches.

  • 13.
    Ghoorchian, Kambiz
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS. SICS Sweden.
    Rahimian, Fatemeh
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Semi-Supervised Multiple Disambiguation2015Ingår i: IEEE Computer Society Conference Publishing Services / [ed] IEEE, IEEE , 2015Konferensbidrag (Refereegranskat)
    Abstract [en]

    Determining the true entity behind an ambiguousword is an NP-Hard problem known as Disambiguation. Previoussolutions often disambiguate a single ambiguous mention acrossmultiple documents. They assume each document contains onlya single ambiguous word and a rich set of unambiguous contextwords. However, nowadays we require fast disambiguation ofshort texts (like news feeds, reviews or Tweets) with few contextwords and multiple ambiguous words. In this research we focuson Multiple Disambiguation (MD) in contrast to Single Disambiguation(SD). Our solution is inspired by a recent algorithm developed for SD. The algorithm categorizes documents by first,transferring them into a graph and then, clustering the graphbased on its topological structure. We changed the graph-baseddocument-modeling of the algorithm, to account for MD. Also,we added a new parameter that controls the resolution of theclustering. Then, we used a supervised sampling approach formerging the clusters when appropriate. Our algorithm, comparedwith the original model, achieved 10% higher quality in termsof F1-Score using only 4% sampling from the dataset.

  • 14.
    Girdzijauskas, Sarunas
    et al.
    EPFL, Switzerland.
    Chockler, Gregory
    IBM Research, Israel.
    Melamed, Roie
    IBM Research, Israel.
    Tock, Yoav
    IBM Research, Israel.
    Gravity: An Interest-Aware Publish/Subscribe System Based on Structured Overlays2008Ingår i: Proceedings of the Second International Conference on Distributed Event-Based Systems, DEBS 2008 / [ed] Roberto Baldoni, 2008Konferensbidrag (Refereegranskat)
  • 15. Girdzijauskas, Sarunas
    et al.
    Chockler, Gregory
    Vifgusson, Ymir
    Tock, Yoav
    Melamed, Roie
    Magnet: Practical Subscription Clustering for Internet-Scale Publish/Subscribe2010Ingår i: DEBS '10 Proceedings of the Fourth ACM International Conference on Distributed Event-Based Systems, 2010, s. 172-183Konferensbidrag (Övrigt vetenskapligt)
    Abstract [en]

    An effective means for building Internet-scale distributed applications, and in particular those involving group-based information sharing, is to deploy peer-to-peer overlay networks. The key pre-requisite for supporting these types of applications on top of the overlays is efficient distribution of messages to multiple subscribers dispersed across numerous multicast groups.

    In this paper, we introduce Magnet: a peer-to-peer publish/subscribe system which achieves efficient message distribution by dynamically organizing peers with similar subscriptions into dissemination structures which preserve locality in the subscription space. Magnet is able to significantly reduce the message propagation costs by taking advantage of subscription correlations present in many large-scale group-based applications.

    We evaluate Magnet by comparing its performance against a strawman pub/sub system which does not cluster similar subscriptions by simulation. We find that Magnet outperforms the strawman by a substantial margin on clustered subscription workloads produced using both generative models and real application traces.

  • 16.
    Girdzijauskas, Sarunas
    et al.
    Ecole Polytechnique Fédérale de Lausanne (EPFL).
    Datta, Anwitaman
    Ecole Polytechnique Fédérale de Lausanne (EPFL).
    Aberer, Karl
    Ecole Polytechnique Fédérale de Lausanne (EPFL).
    On Small World Graphs in Non-uniformly Distributed Key Spaces2005Ingår i: Proceedings - International Workshop on Biomedical Data Engineering, BMDE2005, 2005Konferensbidrag (Refereegranskat)
    Abstract [en]

    In this paper we show that the topologies of most logarithmic-style P2P systems like Pastry, Tapestry or P-Grid resemble small-world graphs. Inspired by Kleinberg's small-world model [7] we extend the model of building "routing-efficient" small-world graphs and propose two new models. We show that the graph, constructed according to our model for uniform key distribution and logarithmic outdegree, will have similar properties as the topologies of structured P2P systems with logarithmic out-degree. Moreover, we propose a novel model of building graphs which support uneven node distributions and preserves all desired properties of Kleinberg's small-world model. With such a model we are setting a reference base for nowadays emerging P2P systems that need to support uneven key distributions

  • 17.
    Girdzijauskas, Sarunas
    et al.
    Ecole Polytech Fed Lausanne.
    Datta, Anwitaman
    Ecole Polytech Fed Lausanne.
    Aberer, Karl
    Ecole Polytech Fed Lausanne.
    Oscar: A Data-Oriented Overlay For Heterogeneous Environments2007Ingår i: 2007 IEEE 23RD INTERNATIONAL CONFERENCE ON DATA ENGINEERING, VOLS 1-3, 2007, s. 1340-1342Konferensbidrag (Refereegranskat)
    Abstract [en]

    Quite a few data-oriented overlay networks have been designed in recent years. These designs often (implicitly) assume various homogeneity which seriously limit their us ability v in real world. In this paper we present some performance results of the Oscar overlay, which simultaneously deals with heterogeneity as observed in the internet (capacity of computers, bandwidth) as well as non-uniformity observed in data-oriented applications.

  • 18.
    Girdzijauskas, Sarunas
    et al.
    Ecole Polytech Fed Lausann.
    Datta, Anwitaman
    Ecole Polytech Fed Lausann.
    Aberer, Karl
    Ecole Polytech Fed Lausann.
    Oscar: Small-world overlay for realistic key distributions2006Ingår i: The Fourth International Workshop on Databases, Information Systems and Peer-to-Peer Computing, September 11, 2006, Seoul, Korea, 2006, s. 247-258Konferensbidrag (Refereegranskat)
    Abstract [en]

    The research on P2P systems which support skewed key distributions has rapidly advanced in the recent years. Yet, the assumptions on the skews we are dealing with remained pretty simple: most of the existing literature assumes simple monotonous key distribution skews. However, this is not always the case. For example, Gnutella filename traces show that complex key-distribut ions rather than monotonous skews occur in practice. We show that one of the seminal P2P systems which support skewed keys - Mercury [7], performs poorly given such complex distributions generated from the trace of Gnutella filenames. We discuss the shortcomings of such state-of-the-art techniques. We present an overlay network Oscar, based on a novel overlay construction mechanism, which does not depend on the key-distribution complexity. We demonstrate through simulations that our technique performs well and significantly surpasses Mercury for such realistic workloads.

  • 19. Girdzijauskas, Sarunas
    et al.
    Datta, Anwitaman
    Aberer, Karl
    Structured Overlay For Heterogeneous Environments: Design and Evaluation of Oscar2010Ingår i: ACM Transactions on Autonomous and Adaptive Systems (TAAS), ISSN 1556-4665, Vol. 5, nr 1Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    Recent years have seen advances in building large Internet-scale index structures, generally known as structured overlays. Early structured overlays realized distributed hash tables (DHTs) which are ill suited for anything but exact queries. The need to support range queries necessitates systems that can handle uneven load distributions. However such systems suffer from practical problems including poor latency, disproportionate bandwidth usage at participating peers, or unrealistic assumptions on peers' homogeneity, in terms of available storage or bandwidth resources. In this article we consider a system that is not only able to support uneven load distributions but also to operate in heterogeneous environments, where each peer can autonomously decide how much of its resources to contribute to the system. We provide the theoretical foundations of realizing such a network and present a newly proposed system Oscar based on these principles. Oscar can construct efficient overlays given arbitrary load distributions by employing a novel scalable network sampling technique. The simulations of our system validate the theory and evaluate Oscar's performance under typical challenges, encountered in real-life large-scale networked systems, including participant heterogeneity, faults, and skewed and dynamic load-distributions. Thus the Oscar distributed index fills in an important gap in the family of structured overlays, bringing into life a practical Internet-scale index, which can play a crucial role in enabling data-oriented applications distributed over wide-area networks. 

  • 20. Girdzijauskas, Sarunas
    et al.
    Galuba, Wojciech
    Darlagiannis, Vasilios
    Datta, Anwitaman
    Aberer, Karl
    Fuzzynet: Ringless Routing in a Ring-like Structured Overlay2010Ingår i: Peer-to-Peer Networking and Applications, ISSN 1936-6442, E-ISSN 1936-6450, Vol. 4, nr 3, s. 259-273Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    Many structured overlay networks rely on a ring invariant as a core network connectivity element. The responsibility ranges of the participating peers and navigability principles (greedy routing) heavily depend on the ring structure. For correctness guarantees, each node needs to eagerly maintain its immediate neighboring links - the ring invariant. However, the ring maintenance is an expensive task and it may not even be possible to maintain the ring invariant continuously under high churn, particularly as the network size grows. Furthermore, routing anomalies in the network, peers behind firewalls and Network Address Translators (NATs) create non-transitivity effects, which inevitably lead to the violation of the ring invariant. We argue that reliance on the ring structure is a serious impediment for real life deployment and scalability of structured overlays. In this paper we propose an overlay called Fuzzynet, which does not rely on the ring invariant, yet has all the functionalities of structured overlays. Fuzzynet takes the idea of lazy overlay maintenance further by dropping any explicit connectivity and data maintenance requirement, relying merely on the actions performed when new Fuzzynet peers join the network. We show that with sufficient amount of neighbors (O(log N), comparable to traditional structured over-lays), even under high churn, data can be retrieved in Fuzzynet w.h.p. We validate our novel design principles by simulations as well as PlanetLab experiments and compare them with ring based overlays.

  • 21. Guerrieri, A.
    et al.
    Rahimian, Fatemeh
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Montresor, A.
    Tovel: Distributed Graph Clustering for Word Sense Disambiguation2017Ingår i: IEEE International Conference on Data Mining Workshops, ICDMW, IEEE Computer Society, 2017, s. 623-630, artikel-id 7836725Konferensbidrag (Refereegranskat)
    Abstract [en]

    Word sense disambiguation is a fundamental problem in natural language processing (NLP). In this problem, a large corpus of documents contains mentions to well-known (non-Ambiguous) words, together with mentions to ambiguous ones. The goal is to compute a clustering of the corpus, such that documents that refer to the same meaning appear in the same cluster, subsequentially, each cluster is assigned to a different semantic meaning. In this paper, we propose a mechanism for word sense disambiguation based on distributed graph clustering that is incremental in nature and can scale to big data. A novel, heuristic vertex-centric algorithm based on the metaphor of the water cycle is used to cluster the graph. Our approach is evaluated on real datasets in both centralized and decentralized environments.

  • 22.
    Kefato, Zekarias
    et al.
    Trento Univesrity.
    Sheikh, Nasrullah
    Trento University.
    Bahri, Leila
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Soliman, Amira
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Montresor, Alberto
    Trento University.
    CAS2VEC: Network-Agnostic Cascade Prediction in Online Social Networks2018Ingår i: The 5th International Symposium on Social Networks Analysis, Management and Security (SNAMS-2018), IEEE, 2018Konferensbidrag (Refereegranskat)
  • 23.
    Kefato, Zekarias
    et al.
    Trento University.
    Sheikh, Nasrullah
    Trento University.
    Bahri, Leila
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Soliman, Amira
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Montresor, Alberto
    Trento University.
    CaTS: Network-Agnostic Virality Prediction Model to Aid Rumour Detection2018Konferensbidrag (Refereegranskat)
  • 24.
    Kefato, Zekarias T.
    et al.
    Univ Trento, Trento, Italy..
    Sheikh, Nasrullah
    Univ Trento, Trento, Italy..
    Bahri, Leila
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Soliman, Amira
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Montresor, Alberto
    Univ Trento, Trento, Italy..
    Girdzijauskas, Sarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    CAS2VEC: Network-Agnostic Cascade Prediction in Online Social Networks2018Konferensbidrag (Refereegranskat)
    Abstract [en]

    Effectively predicting whether a given post or tweet is going to become viral in online social networks is of paramount importance for several applications, such as trend and break-out forecasting. While several attempts towards this end exist, most of the current approaches rely on features extracted from the underlying network structure over which the content spreads. Recent studies have shown, however, that prediction can be effectively performed with very little structural information about the network, or even with no structural information at all. In this study we propose a novel network-agnostic approach called CAS2VEC, that models information cascades as time series and discretizes them using time slices. For the actual prediction task we have adopted a technique from the natural language processing community. The particular choice of the technique is mainly inspired by an empirical observation on the strong similarity between the distribution of discretized values occurrence in cascades and words occurrence in natural language documents. Thus, thanks to such a technique for sentence classification using convolutional neural networks, CAS2VEC can predict whether a cascade is going to become viral or not. We have performed extensive experiments on two widely used real-world datasets for cascade prediction, that demonstrate the effectiveness of our algorithm against strong baselines.

  • 25.
    Khelghatdoust, Mansour
    et al.
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Short: Gossip-based sampling in social overlays2014Ingår i: Networked Systems: Second International Conference, NETYS 2014, Marrakech, Morocco, May 15–17, 2014, Revised Selected Papers, 2014, s. 335-340Konferensbidrag (Refereegranskat)
    Abstract [en]

    Performance of many P2P systems depends on the ability to construct a random overlay network among the nodes. Current state-of-the-art techniques for constructing random overlays have an implicit requirement that any two nodes in the system should always be able to communicate and establish a link between them. However, this is not the case in some of the environments where distributed systems are required to be deployed, e.g., Decentralized Online Social Networks, Wireless networks, or networks with limited connectivity because of NATs/firewalls, etc. In this paper we propose a gossip based peer sampling service capable of running on top of such restricted networks and producing an on-the-fly random overlay. The service provides every participating node with a set of uniform random nodes from the network, as well as efficient routing paths for reaching those nodes via the restricted network.

  • 26.
    Klemm, Fabius
    et al.
    Ecole Polytech. Fed. de Lausanne (EPFL), Lausanne .
    Girdzijauskas, Sarunas
    Ecole Polytech. Fed. de Lausanne (EPFL), Lausanne .
    Le Boudec, J.-Y.
    Ecole Polytech. Fed. de Lausanne (EPFL), Lausanne .
    Aberer, Karl
    Ecole Polytech. Fed. de Lausanne (EPFL), Lausanne .
    On Routing in Distributed Hash Tables2007Ingår i: Peer-to-Peer Computing, 2007. P2P 2007. Seventh IEEE International Conference on, 2007, s. 113-122Konferensbidrag (Refereegranskat)
    Abstract [en]

    There have been many proposals for constructing routing tables for distributed hash tables (DHT). They can be classified into two groups: A) those that assume that the peers are uniformly randomly distributed in the identifier space, and B) those that allow order-preserving hash functions that lead to a skewed peer distribution in the identifier space. Good solutions for group A have been known for many years. However, DHTs in group A are limited to use randomized hashing and therefore, queries over whole identifier ranges thus do not scale. Group B can handle such queries easily. However, it is more difficult to connect the peers such that the resulting topology provides efficient routing, small routing tables, and balanced routing load. We present an elegant new solution to construct an efficient DHT for group B. Our main idea is to decouple the identifier space from the routing topology. In consequence, our DHT allows arbitrarily skewed peer distributions in the identifier space and does not require the overhead of sampling. Furthermore, the table construction is cheap and does not require active replacement of lost routing entries. To evaluate the performance of routing cost and table construction under high churn, we built an efficient simulator. Using the right data structures, we can easily process the state of over one million peers in RAM.

  • 27.
    Laleh, N.
    et al.
    KTH. University of Insubria, Italy.
    Carminati, B.
    KTH. University of Insubria, Italy.
    Ferrari, Elena
    KTH. University of Insubria, Italy.
    Girdzijauskas, Sarunas
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS. University of Insubria, Italy.
    Gossip-based behavioral group identification in decentralized OSNs2016Ingår i: 12th International Conference on Machine Learning and Data Mining in Pattern Recognition, MLDM 2016, Springer, 2016, s. 676-691Konferensbidrag (Refereegranskat)
    Abstract [en]

    DOSNs are distributed systems providing social networking services that become extremely popular in recent years. In DOSNs, the aim is to give the users control over their data and keeping data locally to enhance privacy. Therefore, identifying behavioral groups of users that share the same behavioral patterns in decentralized OSNs is challenging. In the fully distributed social graph, each user has only one feature vector and these vectors can not move to any central storage or other users in a raw form duo to privacy issues. We use a gossip learning approach where all users are involved with their local estimation of the clustering model and improve their estimations and finally converge to a final clustering model available for all users. In order to evaluate our approach, we implement our algorithm and test it in a real Facebook graph.

  • 28.
    Nasir, M. Anis U.
    et al.
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Rahimian, F.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Gossip-based partitioning and replication for Online Social Networks2014Ingår i: ASONAM 2014 - Proceedings of the 2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, 2014, s. 33-42Konferensbidrag (Refereegranskat)
    Abstract [en]

    Online Social Networks (OSNs) have been gaining tremendous growth and popularity in the last decade, as they have been attracting billions of users from all over the world. Such networks generate petabytes of data from the social interactions among their users and create many management and scalability challenges. OSN users share common interests and exhibit strong community structures, which create complex dependability patterns within OSN data, thus, make it difficult to partition and distribute in a data center environment. Existing solutions, such as, distributed databases, key-value stores and auto scaling services use random partitioning to distribute the data across a cluster, which breaks existing dependencies of the OSN data and may generate huge inter-server traffic. Therefore, there is a need for intelligent data allocation strategy that can reduce the network cost for various OSN operations. In this paper, we present a gossip-based partitioning and replication scheme that efficiently splits OSN data and distributes the data across a cluster. We achieve fault tolerance and data locality, for one-hop neighbors, through replication. Our main contribution is a social graph placement strategy that divides the social graph into predefined size partitions and periodically updates the partitions to place socially connected users together. To evaluate our algorithm, we compare it with random partitioning and a state-of-the-art solution SPAR. Results show that our algorithm generates up to four times less replication overhead compared to random partitioning and half the replication overhead compared to SPAR.

  • 29.
    Nasir, Muhammad Anis Uddin
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Gionis, A.
    De Francisci Morales, G.
    Girdzijauskas, Sarunas
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Fully dynamic algorithm for top-k densest subgraphs2017Ingår i: CIKM '17 Proceedings of the 2017 ACM on Conference on Information and Knowledge Management, Association for Computing Machinery (ACM), 2017, s. 1817-1826Konferensbidrag (Refereegranskat)
    Abstract [en]

    Given a large graph, the densest-subgraph problem asks to find a subgraph with maximum average degree. When considering the top-k version of this problem, a nattive solution is to iteratively find the densest subgraph and remove it in each iteration. However, such a solution is impractical due to high processing cost. The problem is further complicated when dealing with dynamic graphs, since adding or removing an edge requires re-running the algorithm. In this paper, we study the top-k densest-subgraph problem in the sliding-window model and propose an efficient fully-dynamic algorithm. The input of our algorithm consists of an edge stream, and the goal is to find the node-disjoint subgraphs that maximize the sum of their densities. In contrast to existing state-of-the-art solutions that require iterating over the entire graph upon any update, our algorithm profits from the observation that updates only affect a limited region of the graph. Therefore, the top-k densest subgraphs are maintained by only applying local updates. We provide a theoretical analysis of the proposed algorithm and show empirically that the algorithm offen generates denser subgraphs than state-of-the-art competitors. Experiments show an improvement in efficiency of up to five orders of magnitude compared to state-of-the-art solutions.

  • 30.
    Nasir, Muhammad Anis Uddin
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Socially-aware distributed hash tables for decentralized online social networks2015Ingår i: Peer-to-Peer Computing (P2P), 2015 IEEE International Conference on, IEEE Press, 2015, s. 1-10Konferensbidrag (Refereegranskat)
    Abstract [en]

    Many decentralized online social networks (DOSNs) have been proposed due to an increase in awareness related to privacy and scalability issues in centralized social networks. Such decentralized networks transfer processing and storage functionalities from the service providers towards the end users. DOSNs require individualistic implementation for services, (i.e., search, information dissemination, storage, and publish/subscribe). However, many of these services mostly perform social queries, where OSN users are interested in accessing information of their friends.

    In our work, we design a socially-aware distributed hash table (DHTs) for efficient implementation of DOSNs. In particular, we propose a gossip-based algorithm to place users in a DHT, while maximizing the social awareness among them. Through a set of experiments, we show that our approach reduces the lookup latency by almost 30% and improves the reliability of the communication by nearly 10% via trusted contacts.

  • 31.
    Nasir, Muhammad Anis Uddin
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Horii, Hiroshi
    Serafini, Marco
    Kourtellis, Nicolas
    Raymond, Rudy
    Girdzijauskas, Sarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Osogami, Takayuki
    Load Balancing for Skewed Streams on Heterogeneous ClustersManuskript (preprint) (Övrigt vetenskapligt)
    Abstract [en]

    Streaming applications frequently encounter skewed workloads and execute on heterogeneous clusters. Optimal re- source utilization in such adverse conditions becomes a challenge, as it requires inferring the resource capacities and input distribution at run time. In this paper, we tackle the aforementioned challenges by modeling them as a load balancing problem. We propose a novel partitioning strategy called Consistent Grouping (CG), which enables each processing element instance (PEI) to process the workload according to its capacity. The main idea behind CG is the notion of small, equal-sized “virtual workers” at the sources, which are assigned to physical workers based on their capacities. We provide a theoretical analysis of the proposed algorithm and show via extensive empirical evaluation that our proposed scheme outperforms the state-of-the-art approaches, like key grouping. In particular, CG achieves 3.44x better performance in terms of latency compared to key grouping.

  • 32.
    Rahimian, Fatemeh
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Haridi, Seif
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Parallel Community Detection For Cross-Document Coreference2014Konferensbidrag (Refereegranskat)
    Abstract [en]

    This paper presents a highly parallel solution for cross-document coreference resolution, which can deal with billions of documents that exist in the current web. At the core of our solution lies a novel algorithm for community detection in large scale graphs. We operate on graphs which we construct by representing documents' keywords as nodes and the co-location of those keywords in a document as edges. We then exploit the particular nature of such graphs where coreferent words are topologically clustered and can be efficiently discovered by our community detection algorithm. The accuracy of our technique is considerably higher than that of the state of the art, while the convergence time is by far shorter. In particular, we increase the accuracy for a baseline dataset by more than 15% compared to the best reported result so far. Moreover, we outperform the best reported result for a dataset provided for the Word Sense Induction task in SemEval 2010.

  • 33.
    Rahimian, Fatemeh
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Payberah, Amir H.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Haridi, Seif
    KTH, Skolan för informations- och kommunikationsteknik (ICT).
    Subscription Awareness Meets Rendezvous Routing2012Konferensbidrag (Refereegranskat)
    Abstract [en]

    Publish/subscribe communication model has become an indispensable part of the Web 2.0 applications, such as social networks and news syndication. Although there exist a few systems that provide a genuinely scalable service for topic-based publish/subscribe model, the content-based solutions are still suffering from restricted subscription schemes, heavy and unbalanced load on the participating nodes, or excessively high matching complexity. We address these problems by constructing a distributed content-based publish/subscribe system by using only those components that are proven to be scalable and can withstand the workloads of massive sizes. Our publish/subscribe solution, Vinifera, requires only a bounded node degree and as we show, through simulations, it scales well to large network sizes and remains efficient under various subscription patterns and loads.

  • 34.
    Rahimian, Fatemeh
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Kommunikation: Infrastruktur och tjänster (Stängd 20120101), Programvaru- och datorsystem, SCS (Stängd 20120101). Swedish Institute of Computer Science (SICS), Stockholm.
    Girdzijauskas, Sarunas
    Swedish Institute of Computer Science (SICS), Stockholm.
    Payberah, Amir H.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Kommunikation: Infrastruktur och tjänster (Stängd 20120101), Programvaru- och datorsystem, SCS (Stängd 20120101). Swedish Institute of Computer Science (SICS), Stockholm.
    Haridi, Seif
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Kommunikation: Infrastruktur och tjänster (Stängd 20120101), Programvaru- och datorsystem, SCS (Stängd 20120101). Swedish Institute of Computer Science (SICS), Stockholm.
    Vitis: A Gossip-based Hybrid Overlay for Internet-scale Publish/Subscribe Enabling Rendezvous Routing in Unstructured Overlay Networks2011Ingår i: Proceedings - 25th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2011, 2011, s. 746-757Konferensbidrag (Refereegranskat)
    Abstract [en]

    Peer-to-peer overlay networks are attractive solutions for building Internet-scale publish/subscribe systems. However, scalability comes with a cost: a message published on a certain topic often needs to traverse a large number of uninterested (unsubscribed) nodes before reaching all its subscribers. This might sharply increase resource consumption for such relay nodes (in terms of bandwidth transmission cost, CPU, etc) and could ultimately lead to rapid deterioration of the system's performance once the relay nodes start dropping the messages or choose to permanently abandon the system. In this paper, we introduce {\em Vitis}, a gossip-based publish/subscribe system that significantly decreases the number of relay messages, and scales to an unbounded number of nodes and topics. This is achieved by the novel approach of enabling rendezvous routing on unstructured overlays. We construct a hybrid system by injecting structure into an otherwise unstructured network. The resulting structure resembles a navigable small-world network, which spans along clusters of nodes that have similar subscriptions. The properties of such an overlay make it an ideal platform for efficient data dissemination in large-scale systems. We perform extensive simulations and evaluate Vitis by comparing its performance against two base-line publish/subscribe systems: one that is oblivious to node subscriptions, and another that exploits the subscription similarities. Our measurements show that Vitis significantly outperforms the base-line solutions on various subscription and churn scenarios, from both synthetic models and real-world traces

  • 35.
    Rahimian, Fatemeh
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Nguyen Huu, Thinh Le
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Locality Awareness in a Peer-to-Peer Publish/Subscribe System2012Ingår i: Distributed applications and interoperable systems: 12th IFIP WG 6.1 International Conference, DAIS 2012, Stockholm, Sweden, June 13-16, 2012. Proceedings, Springer, 2012, Vol. 7272, s. 45-58Konferensbidrag (Refereegranskat)
    Abstract [en]

    Peer-to-peer publish/subscribe systems are promising solutions to provide distributed content distribution services at Internet-scale with low cost. One of the potential problems with peer-to-peer overlays, however, is the inefficient traffic and large delays, due to the mismatch between the physical network and the overlay topology. This paper introduces a locality-aware extension to a peer-to-peer publish/subscribe system, named Vitis. The ultimate purpose is to avoid communications over long-distance links, instead, nodes send data over short-distance and low-cost links, when possible, while maintaining an acceptable quality of service. We show, through simulations, that the average data delivery time is up to 40% improved. The cost to pay is at most 10% more relaying in the peer-to-peer overlay.

  • 36.
    Rahimian, Fatemeh
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Payberah, Amir H.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Haridi, Seif
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Distributed Vertex-Cut Partitioning2014Ingår i: In the 14th IFIP international conference on Distributed Applications and Interoperable Systems (DAIS’14)., 2014, s. 186-200Konferensbidrag (Refereegranskat)
    Abstract [en]

    Graph processing has become an integral part of big data analytics. With the ever increasing size of the graphs, one needs to partition them into smaller clusters, which can be managed and processed more easily on multiple machines in a distributed fashion. While there exist numerous solutions for edge-cut partitioning of graphs, very little effort has been made for vertex-cut partitioning. This is in spite of the fact that vertex-cuts are proved significantly more effective than edge-cuts for processing most real world graphs. In this paper we present Ja-be-Ja-vc, a parallel and distributed algorithm for vertex-cut partitioning of large graphs. In a nutshell, Ja-be-Ja-vc is a local search algorithm that iteratively improves upon an initial random assignment of edges to partitions. We propose several heuristics for this optimization and study their impact on the final partitioning. Moreover, we employ simulated annealing technique to escape local optima. We evaluate our solution on various graphs and with variety of settings, and compare it against two state-of-the-art solutions. We show that Ja-be-Ja-vc outperforms the existing solutions in that it not only creates partitions of any requested size, but also requires a vertex-cut that is better than its counterparts and more than 70% better than random partitioning.

  • 37.
    Rahimian, Fatemeh
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS. Computer System Lab. (CSL), SICS Swedish, Sweden.
    Payberah, Amir H.
    Computer System Lab. (CSL), SICS Swedish, Sweden.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Jelasity, Mark
    MTA SZTE Research Group on AI, Hungarian Academy of Sciences and University of Szeged, Hungary.
    Haridi, Seif
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    A Distributed Algorithm for Large-Scale Graph Partitioning2015Ingår i: ACM Transactions on Autonomous and Adaptive Systems, ISSN 1556-4665, E-ISSN 1556-4703, Vol. 10, nr 2, artikel-id 12Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    Balanced graph partitioning is an NP-complete problem with a wide range of applications. These applications include many large-scale distributed problems, including the optimal storage of large sets of graph-structured data over several hosts. However, in very large-scale distributed scenarios, state-of-the-art algorithms are not directly applicable because they typically involve frequent global operations over the entire graph. In this article, we propose a fully distributed algorithm called JA-BE-JA that uses local search and simulated annealing techniques for two types of graph partitioning: edge-cut partitioning and vertex-cut partitioning. The algorithm is massively parallel: There is no central coordination, each vertex is processed independently, and only the direct neighbors of a vertex and a small subset of random vertices in the graph need to be known locally. Strict synchronization is not required. These features allow JA-BE-JA to be easily adapted to any distributed graph-processing system from data centers to fully distributed networks. We show that the minimal edge-cut value empirically achieved by JA-BE-JA is comparable to state-of-the-art centralized algorithms such as Metis. In particular, on large social networks, JA-BE-JA outperforms Metis. We also show that JA-BE-JA computes very low vertex-cuts, which are proved significantly more effective than edge-cuts for processing most real-world graphs.

  • 38.
    Rahimian, Fatemeh
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Payberah, Amir H.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Jelasity, Mark
    Haridi, Seif
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    JA-BE-JA: A Distributed Algorithm for Balanced Graph Partitioning2013Ingår i: 7th International Conference on Self-Adaptive and Self-Organizing Systems (SASO), 2013 IEEE, IEEE conference proceedings, 2013, s. 51-60Konferensbidrag (Refereegranskat)
    Abstract [en]

    Balanced graph partitioning is a well known NP-complete problem with a wide range of applications. These applications include many large-scale distributed problems including the optimal storage of large sets of graph-structured data over several hosts-A key problem in today's Cloud infrastructure. However, in very large-scale distributed scenarios, state-of-the-Art algorithms are not directly applicable, because they typically involve frequent global operations over the entire graph. In this paper, we propose a fully distributed algorithm, called JA-BE-JA, that uses local search and simulated annealing techniques for graph partitioning. The algorithm is massively parallel: there is no central coordination, each node is processed independently, and only the direct neighbors of the node, and a small subset of random nodes in the graph need to be known locally. Strict synchronization is not required. These features allow JA-BE-JA to be easily adapted to any distributed graph-processing system from data centers to fully distributed networks. We perform a thorough experimental analysis, which shows that the minimal edge-cut value achieved by JA-BE-JA is comparable to state-of-the-Art centralized algorithms such as METIS. In particular, on large social networks JA-BEJA outperforms METIS, which makes JA-BE-JA-A bottom-up, self-organizing algorithm-A highly competitive practical solution for graph partitioning.

  • 39. Sedaghat, M.
    et al.
    Hernandez-Rodriguez, F.
    Elmroth, E.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Divide the task, multiply the outcome: Cooperative VM consolidation2015Ingår i: Proceedings of the International Conference on Cloud Computing Technology and Science, CloudCom, IEEE conference proceedings, 2015, nr February, s. 300-305Konferensbidrag (Refereegranskat)
    Abstract [en]

    Efficient resource utilization is one of the main concerns of cloud providers, as it has a direct impact on energy costs and thus their revenue. Virtual machine (VM) consolidation is one the common techniques, used by infrastructure providers to efficiently utilize their resources. However, when it comes to large-scale infrastructures, consolidation decisions become computationally complex, since VMs are multi-dimensional entities with changing demand and unknown lifetime, and users often overestimate their actual demand. These uncertainties urges the system to take consolidation decisions continuously in a real time manner. In this work, we investigate a decentralized approach for VM consolidation using Peer to Peer (P2P) principles. We investigate the opportunities offered by P2P systems, as scalable and robust management structures, to address VM consolidation concerns. We present a P2P consolidation protocol, considering the dimensionality of resources and dynamicity of the environment. The protocol benefits from concurrency and decentralization of control and it uses a dimension aware decision function for efficient consolidation. We evaluate the protocol through simulation of 100,000 physical machines and 200,000 VM requests. Results demonstrate the potentials and advantages of using a P2P structure to make resource management decisions in large scale data centers. They show that the P2P approach is feasible and scalable and produces resource utilization of 75% when the consolidation aim is 90%.

  • 40.
    Soliman, Amira
    et al.
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Bahri, Leila
    Insubria University, Italy.
    Carminati, Barbara
    Insubria University, Italy.
    Ferrari, Elena
    Insubria University, Italy.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    DIVa: Decentralized Identity Validation for Social Networks2015Rapport (Övrigt vetenskapligt)
    Abstract [en]

    We suggested DIVa, a decentralized, unsupervised, and association rule mining based solution for the learning of fine-grained correlations between profile attributes in Online Social Networks. These correlations can be used for identity validation purposes. In this report, we provide the technical details and the security analysis proofs of the DIVa model.

  • 41.
    Soliman, Amira
    et al.
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Bahri, Leila
    Carminati, Barbara
    Ferrari, Elena
    Girdzijauskas, Sarunas
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    DIVa: Decentralized Identity Validation for Social Networks2015Ingår i: PROCEEDINGS OF THE 2015 IEEE/ACM INTERNATIONAL CONFERENCE ON ADVANCES IN SOCIAL NETWORKS ANALYSIS AND MINING (ASONAM 2015), Association for Computing Machinery (ACM), 2015, s. 383-391Konferensbidrag (Refereegranskat)
    Abstract [en]

    Online Social Networks exploit a lightweight process to identify their users so as to facilitate their fast adoption. However, such convenience comes at the price of making legitimate users subject to different threats created by fake accounts. Therefore, there is a crucial need to empower users with tools helping them in assigning a level of trust to whomever they interact with. To cope with this issue, in this paper we introduce a novel model, DIVa, that leverages on mining techniques to find correlations among user profile attributes. These correlations are discovered not from user population as a whole, but from individual communities, where the correlations are more pronounced. DIVa exploits a decentralized learning approach and ensures privacy preservation as each node in the OSN independently processes its local data and is required to know only its direct neighbors. Extensive experiments using real-world OSN datasets show that DIVa is able to extract fine-grained community-aware correlations among profile attributes with average improvements up to 50% than the global approach.

  • 42.
    Soliman, Amira
    et al.
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Bahri, Leila
    Girdzijauskas, Šarunas
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Carminati, Barbara
    Ferrari, Elena
    CADIVa: Cooperative and Adaptive Decentralized Identity Validation Model for Social Networks2016Ingår i: Social Network Analysis and Mining, ISSN 1869-5450, E-ISSN 1869-5469, Vol. 6, nr 1, artikel-id UNSP 36Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    Online social networks (OSNs) have successfully changed the way people interact. Online interactions among people span geographical boundaries and interweave with different human life activities. However, current OSNs identification schemes lack guarantees on quantifying the trustworthiness of online identities of users joining them. Therefore, driven from the need to empower users with an identity validation scheme, we introduce a novel model, cooperative and adaptive decentralized identity validation CADIVa, that allows OSN users to assign trust levels to whomever they interact with. CADIVa exploits association rule mining approach to extract the identity correlations among profile attributes in every individual community in a social network. CADIVa is a fully decentralized and adaptive model that exploits fully decentralized learning and cooperative approaches not only to preserve users privacy, but also to increase the system reliability and to make it resilient to mono-failure. CADIVa follows the ensemble learning paradigm to preserve users privacy and employs gossip protocols to achieve efficient and low-overhead communication. We provide two different implementation scenarios of CADIVa. Results confirm CADIVa's ability to provide fine-grained community-aware identity validation with average improvement up to 36 and 50 % compared to the semi-centralized or global approaches, respectively.

  • 43.
    Soliman, Amira
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Adagraph: Adaptive graph-based algorithms for spam detection in social networks2017Ingår i: 5th International Conference on Networked Systems, NETYS 2017, Springer Verlag , 2017, s. 338-354Konferensbidrag (Refereegranskat)
    Abstract [en]

    In the past years, researchers developed approaches to detect spam in Online Social Networks (OSNs) such as URL blacklisting, spam traps and even crowdsourcing for manual classification. Although previous work has shown the effectiveness of using statistical learning to detect spam, existing work employs supervised schemes that require labeled training data. In addition to the heavy training cost, it is difficult to obtain a comprehensive source of ground truth for measurement. In contrast to existing work, in this paper we present AdaGraph that is a novel graph-based approach for spam detection. AdaGraph is unsupervised, hence it diminishes the need of labeled training data and training cost. Particularly, AdaGraph effectively detects spam in large-scale OSNs by analyzing user behaviors using graph clustering technique. Moreover, AdaGraph continuously updates detected communities to comply with users dynamic interactions and activities. Extensive experiments using Twitter datasets show that AdaGraph detects spam with accuracy 92.3%. Furthermore, the false positive rate of AdaGraph is less than 0.3% that is less than half of the rate achieved by the state-of-the-art approaches.

  • 44.
    Soliman, Amira
    et al.
    KTH, Skolan för elektro- och systemteknik (EES), Kommunikationsnät.
    Girdzijauskas, Sarunas
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Adaptive Graph-based algorithms for Spam Detection in Social Networks2016Rapport (Övrigt vetenskapligt)
    Abstract [en]

    As Online Social Networks (OSNs) continue to grow in popularity, a spam marketplace has emerged that includes services selling fraudulent accounts, as well as acts as nucleus of spammers who propagate large-scale spam campaigns. In the past years, researchers developed approaches to detect spam such as URL blacklisting, spam traps and even crowdsourcing for manual classification. Although previous work has shown the effectiveness of using statistical learning to detect spam, existing work employs supervised schemes that require labeled training data. In addition to the heavy training cost, it is difficult to obtain a comprehensive source of ground truth for measurement. In contrast to existing work, in this paper we present a novel graph-based approach for spam detection. Our approach is unsupervised, hence it diminishes the need of labeled training data and training cost. Particularly, our approach can effectively detect the spam in large-scale OSNs by analyzing user behaviors using graph clustering technique. Moreover, our approach continuously updates detected communities to comply with dynamic OSNs where interactions and activities are evolving rapidly. Extensive experiments using Twitter datasets show that our approach is able to detect spam with accuracy 92.3\%. Furthermore, our approach has false positive rate that is less than 0.3\% that is less than half of the rate achieved by the state-of-the-art approaches.

  • 45.
    Soliman, Amira
    et al.
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för informations- och kommunikationsteknik (ICT), Programvaruteknik och Datorsystem, SCS.
    DLSAS: Distributed Large-Scale Anti-Spam Framework for Decentralized Online Social Networks2016Ingår i: 2016 IEEE 2ND INTERNATIONAL CONFERENCE ON COLLABORATION AND INTERNET COMPUTING (IEEE CIC), IEEE Press, 2016, s. 363-372Konferensbidrag (Refereegranskat)
    Abstract [en]

    In the last decade, researchers and the open source community have proposed various Decentralized Online Social Networks (DOSNs) that remove dependency on centralized online social network providers to preserve user privacy. However, transitioning from centralized to decentralized environment creates various new set of problems, such as adversarial manipulations. In this paper, we present DLSAS, a novel unsupervised and decentralized anti-spam framework for DOSNs. DLSAS provides decentralized spam detection that is resilient to adversarial attacks. DLSAS typifies massively parallel frameworks and exploits fully decentralized learning and cooperative approaches. Furthermore, DLSAS provides a novel defense mechanism for DOSNs to prevent malicious nodes participating in the system by creating a validation overlay to assess the credibility of the exchanged information among the participating nodes and exclude the misbehaving nodes from the system. Extensive experiments using Twitter datasets confirm not only the DLSAS's capability to detect spam with higher accuracy compared to state-of-the-art approaches, but also the DLSAS's robustness against different adversarial attacks.

  • 46.
    Soliman, Amira
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Rahimian, Fatemeh
    RISE SICS.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Stad: Stateful Diffusion for Linear Time Community Detection2018Ingår i: 38th IEEE International Conference on Distributed Computing Systems, 2018Konferensbidrag (Refereegranskat)
    Abstract [en]

    Community detection is one of the preeminent topics in network analysis. Communities in real-world networks vary in their characteristics, such as their internal cohesion and size. Despite a large variety of methods proposed to detect communities so far, most of existing approaches fall into the category of global approaches. Specifically, these global approaches adapt their detection model focusing on approximating the global structure of the whole network, instead of performing approximation at the communities level. Global techniques tune their parameters to “one size fits all” model, so they are quite successful with extracting communities in homogeneous cases but suffer in heterogeneous community size distributions. In this paper, we present a stateful diffusion approach (Stad) for community detection that employs diffusion. Stad boosts diffusion with a conductance-based function that acts like a tuning parameter to control the diffusion speed. In contrast to existing diffusion mechanisms which operate with global and fixed speed, Stad introduces stateful diffusion to treat every community individually. Particularly, Stad controls the diffusion speed at node level, such that each node determines the diffusion speed associated with every possible community membership independently. Thus, Stad is able to extract communities more accurately in heterogeneous cases by dropping “one size fits all” model. Furthermore, Stad employs a vertex-centric approach which is fully decentralized and highly scalable, and requires no global knowledge. So as, Stad can be successfully applied in distributed environments, such as large-scale graph processing or decentralized machine learning. The results with both real-world and synthetic datasets show that Stad outperforms the state-of-the-art techniques, not only in the community size scale issue but also by achieving higher accuracy that is twice the accuracy achieved by the state-of-the-art techniques.

  • 47.
    Sozinov, Konstantin
    et al.
    KTH.
    Vlassov, Vladimir
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Girdzijauskas, Sarunas
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Programvaruteknik och datorsystem, SCS.
    Human Activity Recognition Using Federated Learning2018Ingår i: 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, s. 1103-1111Konferensbidrag (Refereegranskat)
    Abstract [en]

    State-of-the-art deep learning models for human activity recognition use large amount of sensor data to achieve high accuracy. However, training of such models in a data center using data collected from smart devices leads to high communication costs and possible privacy infringement. In order to mitigate aforementioned issues, federated learning can be employed to train a generic classifier by combining multiple local models trained on data originating from multiple clients. In this work we evaluate federated learning to train a human activity recognition classifier and compare its performance to centralized learning by building two models, namely a deep neural network and a softmax regression trained on both synthetic and real-world datasets. We study communication costs as well as the influence of erroneous clients with corrupted data in federated learning setting. We have found that federated learning for the task of human activity recognition is capable of producing models with slightly worse, but acceptable, accuracy compared to centralized models. In our experiments federated learning achieved an accuracy of up to 89 % compared to 93 % in centralized training for the deep neural network. The global model trained with federated learning on skewed datasets achieves accuracy comparable to centralized learning. Furthermore, we identified an important issue of clients with corrupted data and proposed a federated learning algorithm that identifies and rejects erroneous clients. Lastly, we have identified a trade-off between communication cost and the complexity of a model. We show that more complex models such as deep neural network require more communication in federated learning settings for human activity recognition compared to less complex models, such as multinomial logistic regression.

  • 48.
    Steinle, Mirko
    et al.
    Ecole Polytechnique Federale de Lausanne.
    Aberer, Karl
    Ecole Polytechnique Federale de Lausanne.
    Girdzijauskas, Sarunas
    Ecole Polytechnique Federale de Lausanne.
    Lovis, Christian
    Geneva University Hospitals (HUG).
    Mapping Moving Landscapes by Mining Mountains of Logs: Novel Techniques for Dependency Model Generation2006Ingår i: The 32nd International Conference on Very Large Data Bases, September 12-15, 2006, Seoul, Korea., 2006, s. 1093-1102Konferensbidrag (Refereegranskat)
    Abstract [en]

    Problem diagnosis for distributed systems is usually difficult. Thus, an automated support is needed to identify root causes of encountered problems such as performance lags or inadequate functioning quickly. The many tools and techniques existing today that perform this task rely usually on some dependency model of the system. However, in complex and fast evolving environments it is practically unfeasible to keep such a model up-to-date manually and it has to be created in an automatic manner. For high level objects this is in itself a challenging and less studied task. In this paper, we propose three different approaches to discover dependencies by mining system logs. Our work is inspired by a recently developed data mining algorithm and techniques for collocation extraction from the natural language processing field. We evaluate the techniques in a case study for Geneva University Hospitals (HUG) and perform large-scale experiments on production data. Results show that all techniques are capable of finding useful dependency information with reasonable precision in a real-world environment.

1 - 48 av 48
RefereraExporteraLänk till träfflistan
Permanent länk
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf