S
Soumen Chakrabarti
Researcher at Indian Institute of Technology Bombay
Publications - 208
Citations - 16289
Soumen Chakrabarti is an academic researcher from Indian Institute of Technology Bombay. The author has contributed to research in topics: Ranking (information retrieval) & Web page. The author has an hindex of 55, co-authored 208 publications receiving 15481 citations. Previous affiliations of Soumen Chakrabarti include University of California & Indian Institutes of Technology.
Papers
More filters
Proceedings Article
Joint Bootstrapping of Corpus Annotations and Entity Types
TL;DR: TMI is presented, a bipartite graphical model for joint type-mention inference that shows considerable annotation accuracy improvement and compares with Google’s recent annotations of the same corpus with Freebase entities, and reports considerable improvements within the people domain.
Book ChapterDOI
Multi-task Learning for Target-Dependent Sentiment Classification
TL;DR: The authors proposed a multi-task target-dependent sentiment classification system that is informed by feature representation learned for the related auxiliary task of passage-level sentiment classification, which outperforms state-of-the-art baselines.
Proceedings ArticleDOI
Web-scale entity annotation using MapReduce
TL;DR: This work addresses the new and important task of annotating token spans in billions of Web pages that mention named entities from a large entity catalog such as Wikipedia or Freebase, and designed simple but effective application-specific load estimation and key-splitting methods.
Posted Content
Multi-task Learning for Target-dependent Sentiment Classification
TL;DR: MTTDSC is presented, a multi-task target-dependent sentiment classification system that is informed by feature representation learnt for the related auxiliary task of passage-level sentiment classification, and anecdotal evidence that prior systems can make incorrect target-specific predictions because they miss sentiments expressed by words independent of target is presented.
Posted Content
Features and Aggregators for Web-scale Entity Search
Uma Sawant,Soumen Chakrabarti +1 more
TL;DR: This work explores simple, robust, discriminative ranking algorithms, with informative snippet features and broad families of aggregation functions, and presents a "universal" feature encoding which jointly expresses the perplexity (informativeness) of a query term match and the proximity of the match to the entity mention.