Research on the Influence of Sentiment Polarity of Information Retrieval Tasks on Retrieval Behavior

  • Huang Kun ,
  • Guo Qionghao ,
  • Hao Xijia ,
  • Zhou Chenyu
Expand
  • [1School of Government, Beijing Normal University, Beijing, 100875;2Baolong Foreign Language School of Longcheng High School(Education Group), Guangdong, 518000]

Online published: 2025-05-19

Abstract

[Purpose/significance] It explored the impact of task sentiment polarity on retrieval behavior in the context of information retrieval, providing reference for improving the sentiment response of information retrieval systems.[Method/process] Thirty-six subjects were randomly recruited to finish assigned three positive and negative retrieval task separately. Morae software was applied to record 216 retrieval processes. It analyses the differences in retrieval be⁃haviours under positive and negative information retrieval tasks, information retrieval process emotions, as well as the interactive effects of both emotions on retrieval behaviours. [Result/conclusion] The objective polarity of the informa⁃tion retrieval task and the subjective polarity are generally consistent; users conducted more query interactions under the positive tasks, and more in-depth browsing and viewing under the negative ones. The higher the retrieval effort, the more negative the emotions during the retrieval process.The user experience was significantly affected by the retrieval process emotion. Information retrieval task affective polarity showed significant interaction effects with retrieval process emotion on nine indicators related to query, mouse, viewing, and time, suggesting that the degree to which the sentiment polarity of different tasks promotes and inhibits retrieval behavior is regulated by the emotion of the retrieval process.

Cite this article

Huang Kun , Guo Qionghao , Hao Xijia , Zhou Chenyu . Research on the Influence of Sentiment Polarity of Information Retrieval Tasks on Retrieval Behavior[J]. Information and Documentation Services, 2025 , 46(3) : 94 -101 . DOI: 10.12154/j.qbzlgz.2025.03.010

Outlines

/