Back to Talks

Seminar 3 : Automatic identification of user interests to improve personalized search
Speaker : Vatsheel Singh
Date : 12:00 noon, Friday, October 9th, 2009
Venue : CS103

Abstract: With the amount of information available on the web "one size fits all" approach in returning search results is not sustainable.Personalization systems are necessary to provide personalized information access. However users of these systems may not be interested in giving explicit feedback. In this talk we look at various approaches to construct user profiles namely, Topic Sensitive Page Rank, weighted keywords and weighted concepts which use implicit feedback to personalize search.

References :
  1. Talk Slides, available here.
  2. Feng Qiu and Junghoo Cho, Automatic identification of user interest for personalized search, WWW, pages 727-736, 2006.

Seminar 2 : Graph clustering
Speaker : Shashank Srivastav
Date : 12:00 noon, Friday, September 25th, 2009
Venue : CS102

Abstract: In this survey we overview the definitions and methods for graph clustering, that is, finding sets of "related" vertices in graphs. We review the many definitions for what is a cluster in a graph and measures of cluster quality. Then we present global algorithms for producing a clustering for the entire vertex set of an input graph, after which we discuss the task of identifying a cluster for a specific seed vertex by local computation. Some ideas on the application areas of graph clustering algorithms are given. We also address the problematics of evaluating clusterings and benchmarking cluster algorithms.
Paper available from Computer Science Review, Vol 1, No. 1, 2007, p27-64, Graph Clustering by Satu Elisa Schaeffer at http://www.sciencedirect.com/

References :
  1. Talk Slides, available here.
  2. Satu Elisa Schaeffer, Graph clustering, Computer Science Review 1(1): pages 27-64, 2007.

Seminar 1 : Language Learning in Artificial Agents: From Lexicon to Compositionality
Speaker : Ajin George Joseph
Date : 5:00 pm, Tuesday, September 15th, 2009
Venue : CS102

Abstract : The emergence of language is a breakthrough event in the evolution of humans. It revolutionized the way we think, perceive and interact. The study of language origin and evolution is the study of how language emerges from a situation in which there is no language. The problem of language origin and evolution has puzzled the scientific community for a very long period. In this thesis, using simulation modeling, we have tried to investigate this problem. We have developed a model in which a coordinated communication system emerges among a group of agents having cognitive abilities. We have conducted four experiments each of varying complexity. In all the experiment a coherent language emerges among the population of agents. We observed that the nature of the language that emerges in each experiment is different from the other in terms of synonymy, ambiguity and homonymy. We have extended the model to investigate the evolutionary aspect as well. By adding constraints on the cognitive abilities of the agents new properties such as compositionality emerged in the language.

References :
  1. Ajin Joseph George, Language Learning in Artificial Agents: From Lexicon to Compositionality, M.Tech. Dissertation, Indian Institute of Technology Kanpur, 2009. Cached copy available here.
Back to Talks