A Comprehensive Technique for User Activity Based Twitter Content Summarization
Ayushi Gupta1, Devyani Keskar2, Madhur Firodiya3, Siddhi Hagawane4
1Ayushi Gupta*, Student,Dept. Of Computer Engg., MITCOE, Pune.
2Devyani Keskar, Student,Dept. Of Computer Engg., MITCOE, Pune.
3Madhur Firodiya, Student,Dept. Of Computer Engg., MITCOE, Pune.
4Siddhi Hagawane, Student,Dept. Of Computer Engg., MITCOE, Pune.
Manuscript received on March 28, 2020. | Revised Manuscript received on April 25, 2020. | Manuscript published on April 30, 2020. | PP: 1640-1644 | Volume-9 Issue-4, April 2020. | Retrieval Number: D7827049420/2020©BEIESP | DOI: 10.35940/ijeat.D7827.049420
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Going through thousands of comments in order to understand opinion of people on a particular post ingests in a lot of time and resources of the user. By developing this system, we aim that user gets updated with summarized information of all such events in a time constrained manner. It involves merging multiple opinions stated on the social platform and summarizing it to provide the gist of the topic in order to improve ergonomic experience. For this purpose, our system displays both abstractive and extractive summary of the content. Extractive summary generation makes use of Page rank algorithm and abstractive summary generation makes use of RNN (LSTM).
Keywords: Text analysis, Tweets, Live streaming, Filter based analysis, Anomaly detection