Abstract: This paper addresses the issue of keyword extraction from conversations, with the objective of utilizing these watchwords to recover, for every short discussion piece, a little number of conceivably pertinent reports, which can be prescribed to members. In any case, even a short piece contains a mixed bag of words, which are conceivably identified with a few themes; also, utilizing a programmed discourse acknowledgment (ASR) framework presents slips among them. Along these lines, it is hard to surmise correctly the data needs of the discussion members. We first propose a calculation to remove decisive words from the yield of an ASR framework (or a manual transcript for testing), which makes utilization of theme demonstrating methods and of a sub modular prize capacity which supports differing qualities in the magic word set, to coordinate the potential differing qualities of subjects and decrease ASR commotion. At that point, we propose a technique to infer various topically isolated inquiries from this decisive word set, keeping in mind the end goal to amplify the possibilities of making at any rate one pertinent proposal when utilizing these questions to seek over the English Wikipedia. The proposed systems are assessed as far as significance as for discussion pieces from the Fisher, AMI, and ELEA conversational corpora, appraised by a few human judges. The scores demonstrate that our proposition moves forward over past systems that consider just word recurrence or theme closeness, and speaks to a promising answer for a report recommender framework to be utilized as a part of discussions.
Keywords: Document recommendation, information retrieval, keyword extraction, meeting analysis, topic modelling.
Title: Extraction of Keywords from Conversation for Recommending Document
Author: Hitesh Purandare, Abhishek Singh, Niket Vyas, Chhagan Chavan
International Journal of Computer Science and Information Technology Research
ISSN 2348-1196 (print), ISSN 2348-120X (online)
Research Publish Journals
