Abstract

Facts are descriptive user reviews on a specific entity. A number of facts, which contain indirect opinions are called implicit opinions. This study aims to find the word opinion that is implied in the opinion sentence using the help of aspect extraction. Aspect extraction is connected to the opinion word list, thus reducing the search area to find the opinion word. This study had two stages, namely the learning stage and the validation stage. The learning stage focused on co-occurrence matrix opinion extraction using an opinion co-occurrence algorithm and extraction of category of aspects and opinions. The validation phase is to find opinion words that match the implicit opinion sentences based on aspect extraction. The data set is data from hotel user reviews in Yogyakarta from the TripAdvisor website from 2019. The validation results of implicit opinion extraction based on aspect extraction show an accuracy result in aspect extraction of 83.87%, while the accuracy result in opinion extraction is 56.13%.

Original languageEnglish
Title of host publication2022 5th International Seminar on Research of Information Technology and Intelligent Systems, ISRITI 2022
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages781-786
Number of pages6
ISBN (Electronic)9781665455121
DOIs
Publication statusPublished - 2022
Event5th International Seminar on Research of Information Technology and Intelligent Systems, ISRITI 2022 - Virtual, Online, Indonesia
Duration: 8 Dec 20229 Dec 2022

Publication series

Name2022 5th International Seminar on Research of Information Technology and Intelligent Systems, ISRITI 2022

Conference

Conference5th International Seminar on Research of Information Technology and Intelligent Systems, ISRITI 2022
Country/TerritoryIndonesia
CityVirtual, Online
Period8/12/229/12/22

Keywords

  • aspect extraction
  • aspects category and opinions
  • fact
  • implicit opinion
  • opinion co-occurrence algorithm

Fingerprint

Dive into the research topics of 'Aspect-based Extraction of Implicit Opinions using Opinion Co-occurrence Algorithm'. Together they form a unique fingerprint.

Cite this