View based review exploration

Sailaja Rajanala, Manish Singh

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearch

Abstract

In an e-commerce site, reviews help augment basic product descriptions with detailed user experiences and opinions on the product. These reviews greatly influence the purchase decisions of oncoming customers. Since these reviews are haphazardly presented, they call for befitting review exploration techniques. We propose a novel view based review exploration system that can be used to control the depth and content of the retrieved reviews. We propose three views depending on the depth personified by them: the Bird-eye view, the Unidirectional view, and the Microscopic view. The views present a ranked list of relevant reviews that meet the view criteria. The proposed view based exploration recorded precision as high as 0.85.

Original languageEnglish
Title of host publicationCODS-COMAD 2021 - Proceedings of the 3rd ACM India Joint International Conference on Data Science and Management of Data (8th ACM IKDD CODS and 26th COMAD)
EditorsBalaji Vasan Srinivasan, Yogesh Simmhan
Place of PublicationMew York NY USA
PublisherAssociation for Computing Machinery (ACM)
Pages218-222
Number of pages5
ISBN (Electronic)9781450388177
DOIs
Publication statusPublished - 2020
Externally publishedYes
EventACM India Joint International Conference on Data Science and Management of Data 2021 - Online, India
Duration: 2 Jan 20214 Jan 2021
Conference number: 3rd
https://dl.acm.org/doi/proceedings/10.1145/3430984 (Proceedings)
https://cods-comad.in/2021/ (Website)

Conference

ConferenceACM India Joint International Conference on Data Science and Management of Data 2021
Abbreviated titleCODS-COMAD 2021
Country/TerritoryIndia
Period2/01/214/01/21
Internet address

Keywords

  • Information Search and Retrieval
  • Learning to Rank
  • Review Exploration

Cite this