Construction and Consumption of FAQ Knowledge Graphs in the E-Commerce Domain

Amanda C. Ashardian, Fariz Darari, Douglas R. Faisal, Hadi S. Putra

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Frequently Asked Questions (FAQs) are documents or webpages that give answers to common questions a user may ask about a certain topic. This paper presents the construction of a knowledge graph (KG) using data from the FAQ webpages of Tokopedia, one of the largest e-commerce companies in Indonesia. Our approach is based on a top-down method in KG construction, which incorporates web scraping, part-of-speech tagging, verb ontology generation, and Resource Description Framework (RDF)-based KG generation. The KG created in this research has 14795 triples, containing 991 FAQs from 78 categories and 149 subcategories. The findings reveal new consumption capabilities, such as structured queries to FAQs based on verbs and their synonyms, as well as insights gained by exploratory data analysis (EDA).

Original languageEnglish
Title of host publication2023 6th International Conference on Information and Communications Technology, ICOIACT 2023
EditorsAkhmad Dahlan, Yoga Pristyanto, Rifda Faticha Alfa Aziza
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages132-137
Number of pages6
ISBN (Electronic)9798350315639
DOIs
Publication statusPublished - 2023
Event6th International Conference on Information and Communications Technology, ICOIACT 2023 - Yogyakarta, Indonesia
Duration: 10 Nov 2023 → …

Publication series

Name2023 6th International Conference on Information and Communications Technology, ICOIACT 2023

Conference

Conference6th International Conference on Information and Communications Technology, ICOIACT 2023
Country/TerritoryIndonesia
CityYogyakarta
Period10/11/23 → …

Keywords

  • E-Commerce
  • FAQ
  • Knowledge Graph
  • Verb Relations

Fingerprint

Dive into the research topics of 'Construction and Consumption of FAQ Knowledge Graphs in the E-Commerce Domain'. Together they form a unique fingerprint.

Cite this