Collaborative query reformulation in tourism information search
| Date | 11 November 2019 |
| Published date | 11 November 2019 |
| Pages | 1115-1135 |
| DOI | https://doi.org/10.1108/OIR-12-2018-0371 |
| Author | Jia Tina Du,Abu Shamim Mohammad Arif,Preben Hansen |
| Subject Matter | Library & information science,Information behaviour & retrieval,Collection building & management,Bibliometrics,Databases,Information & knowledge management,Information & communications technology,Internet,Records management & preservation,Document management |
Collaborative query reformulation
in tourism information search
Jia Tina Du
School of Information Technology and Mathematical Sciences,
Mawson Lakes Campus, University of South Australia, Adelaide, Australia
Abu Shamim Mohammad Arif
Department of Computer Science and Engineering Discipline,
Khulna University, Khulna, Bangladesh, and
Preben Hansen
Department of Computer and Systems Sciences,
Stockholm University, Stockholm, Sweden
Abstract
Purpose –Collaborative information search (CIS) is a growing and significant research area. Query
formulation and reformulation is an important search strategy in information search. However, limited
research has investigated query behavior during CIS. The purpose of this paper is to characterize
collaborative query reformulation (CQR) by exploring the sources of collaborative query (CQ) terms and the
types and patterns of CQR in the context of tourism information search.
Design/methodology/approach –An empirical study was designed to investigate search query
reformulation as tourists performed CIS on a devised interface. A total of 36 participants (in 18 pairs) took
part in the study; data were documented in pre- and post-search questionnaires, search logs and chat logs.
Findings –The findings show that participants intermixed individual search and collaborative search
during CIS. Participants constructed CQ terms mainly by selecting terms from individual search queries and
discussion chat logs. Eight types of CQR were identified, with specialization (82 percent) accounting for the
most used search tactics. At most times, participants were found to add terms to the previous query. Findings
demonstrated 27 specific CQR patterns; in excess of two-third participants (69 percent) took only one move to
reformulate CQ by adding terms, or replacing/using new words.
Practical implications –The results of this research can be used to inform the design of search systems
supporting collaborative querying in CIS.
Originality/value –This study is highlighting an important research direction of CQ reformulation in
collaborative search while previous studies of the topic are limited, comparing to the vast body of work on
query reformulation in individual information search using regular search systems.
Keywords Information search, Collaborative information search, Collaborative query
Paper type Research paper
Introduction
When a user interacts with an information retrieval system, they usually formulate initial
queries and reformulate these queries during the course of a search (Wildemuth et al., 2018).
A query consists of one or more search terms, and possibly includes logical operators and
modifiers ( Jansen et al., 2000, p. 211). Query formulation involves searchers constructing a
query in the search system to articulate an information need or search intention. Query
reformulation is a key search strategy that users adopt to modify a previous search query in
terms of content or format in order to obtain better search results ( Jansen et al., 2009; Jiang
et al., 2011; Rieh and Xie, 2006). Users have been found to interact with a search system by
frequently reformulating their queries to satisfy their information needs (Aloteibi and
Sanderson, 2014; Jansen et al., 2000; Jansen et al., 2009).
Collaborative information search (CIS) has attracted growing attention in recent years
(Capra et al., 2013; Leeder and Shah, 2016; Wu et al., 2018). Collaborative search refers to an
information searching and retrieving process involving human beings interacting with
Online Information Review
Vol. 43 No. 7, 2019
pp. 1115-1135
© Emerald PublishingLimited
1468-4527
DOI 10.1108/OIR-12-2018-0371
Received 3 December 2018
Revised 7 March 2019
19 June 2019
3 September 2019
Accepted 12 September 2019
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1468-4527.htm
1115
CQR in tourism
information
search
other humans, directly or indirectly, for solving a common task in workplace settings or
everyday life situations (Hansen and Järvelin, 2005). People are often involved in
collaborative search due to the complex and exploratory nature of tasks (Fidel et al., 2004;
Shah, 2013). For example, friends may search together for planning a group trip (Imazu
et al., 2011; Morris, 2008). Tourists have been observed often travel in groups with family or
friends and they tend to perform collaborative search to meet individual needs and
preferences (Fardous et al., 2017; Mohammad Arif et al., 2012). Not only query formulation
and reformulation as important search behaviors in regular (individual) search but they are
an emerging topic in collaborative search (Al-Sammarraie and Umapathy, 2016; Mohammad
Arif et al., 2014; Yue et al., 2012; Yue et al., 2014). For example, Goh et al. (2005) believed that
a user may take advantage of previous/recommended queries and used the appropriate ones
to meet their information need. The authors proposed the concept of collaborative querying
with the aim to assist users in formulating queries to meet information needs by harnessing
other users’expert knowledge or search experience (p. 378). Hust et al. (2002) argued that a
quality information retrieval system should be able to learn from many queries issued from
various users. Hansen and Järvelin (2005) suggested that users could reuse and share initial,
baseline search paths and query construction sequences. Previously learned queries and
their relevant documents have been re-used in new and similar search sessions initiated by
different users to improve the overall retrieval quality for all. Mohammad Arif et al. (2014)
explored the interplay between querying in individual search, communicating between
group members, and querying in collaborative search to achieving the process of CIS in
tourism. In their study, a collaborative query (CQ) refers to the query that is constructed by
more than one user in a group and collaborative query reformulation (CQR) refers to the
process of refining queries by multiple group members during which they add, delete or
modify query terms in order to fulfill their common information needs. The refinement of
queries by group members can be done in the same time (synchronously) or at different
times (asynchronously). In the current study, we adopt the terms CQ and CQR as defined by
Mohammad Arif et al. (2014).
In examining the querying strategy in collaborative search, the above studies shared the
same assumption that query reformulation is based on learned experience and can be
benefitted from previous search or knowledge issued by other users. Whereas being
different from Goh et al.’s (2005) and Hust et al.’s (2002) global query reformation which has
been learned as a list of previously used query formulations from previous search sessions
and then used as suggestions for other users, Mohammad Arif et al. (2014) studied local
collaborative querying that is produced during the execution of collaborative search in
which multiple users searched together as a group explicitly and purposefully toward the
completion of a common task or the achievement of a shared understanding. Researchers
have claimed that local CQR, influences and is influenced by a group of users’search
behavior during CIS (Yue et al., 2012; Yue et al., 2014). Understanding users’querying
patterns can help make automatic query suggestions for a group of users when they
use collaborative search systems (Al-Sammarraie and Umapathy, 2016). Existing studies
have provided insights in collaborative web search query analysis. Nevertheless,
collaborative search has yet been the focus of query reformulation research. We have a
limited understanding of how a group of people interacts with each other to construct
queries together and modify each other’s query.
In this paper, we empirically investigate collaborative properties of query reformulation
in tourism information searching. Studying CQR will contribute to a better understanding of
the CIS process (Mohammad Arif et al., 2014). We assume that, in the course of a
collaborative search, group members not only talk to each other to divide their tasks and
workloads but communicate periodically to discuss search tactics on query reformulation
and share search results. During the search, group members may borrow each other’s ideas,
1116
OIR
43,7
To continue reading
Request your trial