BIS 2009
Paper index 1st BIS 2nd BIS 3rd BIS 4th BIS 5th BIS 6th BIS 7th BIS 8th BIS 9th BIS 10th BIS 11th BIS 12th BIS 13th BIS NEW

 Home
 News
 
 Call for papers
 Call for tutorials
 Call for workshop
 Important dates
 Submission guidelines
 Submission system (external)
 
 Programme
 Keynote speakers
 Tutorials
 Workshops
 FP7 Proposer Session
 
 Proceedings
 Accepted papers
 Best paper award
 Photos NEW
 
 Conference venue
 Fees & registration
 Accommodation
 
 Program committee
 Organizing committee
 
 RSS feed
 

Tutorial on Graph-based Knowledge Discovery in the web and Social Knowledge Management


tutorial at
12th International Conference on Business Information Systems (BIS 2009)

Poznan, Poland



Proliferation of Web 2.0 and Web 3.0 creates massive computer mediated networks. It is expected that by the year 2010, most of the Web information will be created automatically as logs files of Web services (such as Facebook) and by "The Internet of Things". Graphs serve as suitable models for such multidimensional networks where human generated content is an integral part of larger networks which include people and the things they create and do. This tutorial will present novel methods of graph-based mining of large volumes of information. The tutorial will cover applications of these methods to mining multidimensional networks, and we will demonstrate how to make these methods aware of dimensions of networks where people are involved, including social, semantics, and activity management dimensions.

Social networks and social knowledge and ideas management systems are usually characterized by the complex network structures and rich accompanying contextual information. "Explicit" social relations might be augmented with implicit ones extracted from digital traces in social software or from the locational context of people's interactions provided by sensors in physical world.

Avenues to deep socio-semantic analytics and the possibility of high-quality functionalities for socio-technical systems (like recommending people to invite into your social network) hinge on the availability of engines which are able to provide hidden knowledge discovery (like discovering a new relation in a network - that based on the strength of multiple connectivity between the nodes of a social network one can conclude that Dr. Jekyll is related to Mr. Hide), and provide ad hoc generalisation across dimensions. For instance, the ability to detect that a particular person might serve as an representative of a community or as an expert on a particular topic (the example of such generalisation is the expression frequently attributed to Louis XIV "L'e'tat s'est moi (I'm the State).")

Graphs provide a powerful abstraction of the structure and dynamics of diverse kinds of interpersonal or people-to-technology interactions. Through our tutorial we'll demonstrate how various graph-mining techniques (including soft clustering and fuzzy inferencing) can be used to reveal implicit social structures (like to detect ad hoc communities of common interests or practice).

We will outline how to use explicit and implicit social structures to improve the consumability of social knowledge and ideas management web services (such as IdeaJam.com, where people put forward, discuss and rate proposals) by introducing computational models of personalised and contextualised trust, reputation and ranking.

Structure of the tutorial:

  • Introduction to Spreading Activation Methods Framework (SAM Framework)

  • Composition of Multidimensional Networks and Pertaining Navigation Methods

  • Spreading Activation Methods Framework (SAM Framework)

  • Graph-based Algorithms for Ranking and Clustering Digital Contents

  • Main definitions and motivation

  • Main graph-based ranking and clustering algorithms in data mining

  • The Major-Clust algorithm and its fuzzy modification

  • Applications of SAM Framework

  • Optimizing SAM Results

  • Machine-based Localization of Hidden Knowledge in the Internet

  • Social Knowledge Management



Presenter(s)



  • Dr. Alexander Troussov is chief scientist in IBM Ireland Centre for Advanced Studies (CAS) and chief scientist of IBM LanguageWare group. He has published more than 30 peer reviewed journal and conference papers and has 5 patents. In 2000 he joined IBM as the Architect of IBM Dictionary and Linguistic tools group, known now as IBM LanguageWare group. As CAS Chief Scientist, Dr. Alexander Troussov leads IBM Ireland's participation in the 3 year integrated 6th framework EU project NEPOMUK, and is one of the creators of IBM LanguageWare Miner for Multidimensional Socio-Semantic Networks ([here]).


  • Prof. Eugene Levner is Professor of Computer Science at Holon Institute of Technology and Bar-Ilan University, Israel. His main scientific interests are design of computer algorithms, optimization theory, and clustering and classification of digital content. He is author/co-author of seven books and more than 100 articles in refereed journals. His Citation Index is 410, and h-index is 15. He is the full member of the International Academy of Information Sciences, a member of editorial boards of four international journals.


Organizers
12th International Conference on Business Information Systems (BIS 2009), Poznan, Poland 27-29 April 2009
Department of Information Systems, Poznan University of Economics, al. Niepodleglosci 10, 60-967 Poznan, Poland
phone: +48618543632 , fax: +48618543633 , Web: http://bis.kie.ae.poznan.pl/