Applying a Vector Search Method in Reference Service Question-Answer Retrieval Systems

Te Lun Yang*, Guan Lun Huang, Yuen Hsien Tseng

*Corresponding author for this work

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


We have developed a reference service question-answer retrieval system prototype based on dense vector search technology. When librarians receive reference questions, they can utilize reference service question-answer systems to determine if similar questions have been asked before and how they were resolved. Readers can also search the system to see if their questions have been answered. However, traditional reference service question-answer retrieval systems primarily rely on keyword-based searches, which may not capture questions with different phrasing but identical meaning. Vector search has emerged as a promising approach to address this limitation, and it has been widely researched and applied in various domains. Nevertheless, there is a lack of research on applying vector search in reference service question-answer retrieval systems. Thus, we conducted initial experiments in this area. Leveraging ElasticSearch vector search technology, which can efficiently search through large datasets, we adopted this technology in our study. For building our system, we collected question-answer pairs from the Taipei Public Library’s Online Reference Service Platform, one of Taiwan’s most prominent online reference services. We report the data set we used and the system architecture.

Original languageEnglish
Title of host publicationLeveraging Generative Intelligence in Digital Libraries
Subtitle of host publicationTowards Human-Machine Collaboration - 25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023, Proceedings
EditorsDion H. Goh, Shu-Jiun Chen, Suppawong Tuarob
PublisherSpringer Science and Business Media Deutschland GmbH
Number of pages6
ISBN (Print)9789819980840
Publication statusPublished - 2023
Event25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023 - Taipei, Taiwan
Duration: 2023 Dec 42023 Dec 7

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume14457 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023


  • Dense vector search
  • Full-text search engine
  • Question-answer system
  • Reference service
  • Taipei public library

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of 'Applying a Vector Search Method in Reference Service Question-Answer Retrieval Systems'. Together they form a unique fingerprint.

Cite this