Conditioning Customers' Product Reviews for Accurate Classification Performance

Dorothy Yao, Ishani Chatterjee, Meng Chu Zhou

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

Abstract

In recent years, people use Internet as a platform to express their own ideas and opinions about various subjects or products. The data from these sites serve as sources for sentiment analysis. On e-commerce websites, the costumer product review conventionally expresses sentiment that corresponds with the given star rating; however, this is not always true; there are reviews that express sentiments opposite to the given star rating, which can be labeled as outliers. This paper builds on previous work that finds outliers in product review datasets, scraped from Amazon.com, using a statistics-based outlier detection and correction method (SODCM). This work focuses on 3-star reviews specifically and studies the correct polarity assignment of 3-star reviews. It investigates the behavior of SODCM when 3-star reviews are classified as negative and positive respectively.

Original languageEnglish (US)
Title of host publicationICNSC 2022 - Proceedings of 2022 IEEE International Conference on Networking, Sensing and Control
Subtitle of host publicationAutonomous Intelligent Systems
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781665472432
DOIs
StatePublished - 2022
Event19th IEEE International Conference on Networking, Sensing and Control, ICNSC 2022 - Shanghai, China
Duration: Dec 15 2022Dec 18 2022

Publication series

NameICNSC 2022 - Proceedings of 2022 IEEE International Conference on Networking, Sensing and Control: Autonomous Intelligent Systems

Conference

Conference19th IEEE International Conference on Networking, Sensing and Control, ICNSC 2022
Country/TerritoryChina
CityShanghai
Period12/15/2212/18/22

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Computer Networks and Communications
  • Computer Science Applications
  • Control and Optimization

Keywords

  • SODCM
  • Sentiment analysis
  • big data analysis
  • customer review
  • logistic regression
  • machine learning

Fingerprint

Dive into the research topics of 'Conditioning Customers' Product Reviews for Accurate Classification Performance'. Together they form a unique fingerprint.

Cite this