My Work

Blogging is sometimes viewed as a new, grassroots form of journalism and a way to shape democracy outside the mass media and conventional party politics.Blog sites devoted to politics and punditry, as well as to sharing technical developments,receive thousands of hits a day. But the vast majority of blogs are written by ordinary people for much smaller audiences.These blogs are used as a diary or daily journal, a community forum,a comentary to express opinion, a cathersis to let out one's thoughts and feelings; the motivation for writing blogs is varied.

Since there is so much variation in blog genre and style (originating from the variation in the writers motivation), more than topic-related search will be required to satisfy the user. Faceted search enables us to look at the data from its different facets or faces. For blogs, faceted retrieval is not limited to topics but retrieves blogs based on their opiniatedness, genre, writing style and many other.

My work consists of a baseline retrieval subsystem which performs the baseline blog distillation according to topic,an opinion identification subsystem,an in-depth analysis subsystem and personal blog identification subsystem which performs the faceted blog distillation task. In the baseline system, documents which are deemed relevant are retrieved by the retrieval system with respect to the query, without taking into consideration of any facet requirements.A new weighting model is introduced in the baseline retrieval system which weighs those words highly which are present in lesser number of documents with greater density.In the opinionated vs. factual and personal vs. official faceted retrieval,the results obtained in baseline retrival is postprocessed based on features selected and then scored and ranked according to the facet.In the in-depth vs. shallow faceted task, the depth of the text is measured by coherence and coverage of the topic and then these are scored and ranked according to score.