Open AccessSupporting exploratory search tasks with interactive user modeling
Author(s) -
Ruotsalo Tuukka,
Athukorala Kumaripaba,
Głowacka Dorota,
Konyushkova Ksenia,
Oulasvirta Antti,
Kaipiainen Samuli,
Kaski Samuel,
Jacucci Giulio
Publication year - 2013
Publication title -
proceedings of the american society for information science and technology
Language(s) - English
Resource type - Journals
eISSN - 1550-8390
pISSN - 0044-7870
DOI - 10.1002/meet.14505001040
Subject(s) - computer science , exploratory search , relevance (law) , user modeling , human–computer interaction , task (project management) , relevance feedback , exploratory research , domain (mathematical analysis) , user information , information retrieval , baseline (sea) , user interface , user experience design , exploratory analysis , world wide web , information system , data science , artificial intelligence , mathematics , anthropology , law , image (mathematics) , mathematical analysis , oceanography , sociology , engineering , operating system , management , political science , electrical engineering , economics , geology , image retrieval
This paper presents the design and study of interactive user modeling to support exploratory search tasks. Contrary to traditional interactions, such as query based search, query suggestions, or relevance feedback, interactive user modeling allows a user to perceive the state of the user model at all times and provide feedback that directly rewards or penalizes it. The technique allows the user to continuously tune the system's belief about the user's evolving information needs. We demonstrate that such functionality is useful in exploratory search where users need to get accustomed to a body of literature in a domain. We conducted two experiments where scientists carried out exploratory search tasks with our implementation of an interactive user modeling and retrieval system (SciNet) and two baselines: SciNet from which interactive user modeling was excluded and a real‐world baseline (Google Scholar). The results show that interactive user modeling can help users to more effectively find relevant, novel and diverse information without compromises in task execution time.