Page 271 - JOURNAL OF LIBRARY SCIENCE IN CHINA 2015 Vol. 41

270 Journal of Library Science in China, Vol. 7, 2015

Language network of text has its unique characteristics. Various relations exist between terms,
for example, co-occurrence relation, syntactic relation and semantic relation. How to construct
complex text network which can reveal the relations of terms at the same time is one of the
research directions in the future. Further studies need to address how to identify cohesive subgraph
in complex text network according to relations between terms and topological structure of graph.
In addition, the measure according to which these subgraphs are clustered to reveal core sub-
themes and the relations of themes in texts also needs to be discussed.

A review on information search behavior under a multi-device environment

1 ∗
Dan WU & Shaobo LIANG
With the rapid development of the mobile internet, users can employ smarter devices to find
information, and it is increasingly popular to search across multiple devices. This paper examines
users’ search behaviors in multi-devices environment. We utilized literature survey method and
retrieved relevant articles using “multi-device search”, “cross-device search”, “cross-platform
search”, “multi screens search”, “cross-screen search” and “cross sessions search” as search
terms to search in ACM Digital Library, Taylor & Francis, Springer, Emerald, WOS, EBSCO,
Google scholar, CNKI, Baidu and Google. We identified a total of 72 highly relevant articles. The
classical grounded theory method was used to encode the original contents of the references via
ATLAS.ti. These original contents of 72 references include experimental data analysis results, user
interview or questionnaire survey results, user behavior research conclusions, etc.
The results of our study show that users’ search behavior in the multi-device environment has
the following characteristics: 1) The users’ information needs are more diverse and rich, and their
search processes are more complex since they may need more search sessions to complete the
task. 2) Users’ information search behaviors are influenced by the search content, search time and
search location. 3) Interruptions to users’ search may come from external environment, function of
the device and other factors, which increase the chance of multi-device search. Especially, when
users can not satisfy their complex information needs in long search sessions or through several
simple queries, their search processes often involve multiple devices. 4) Device transition may be
influenced by the device, search task, search process, and environmental factors. One important
future task is to deepen theoretical research and constructing the relevant search behavior models.

*Correspondence should be addressed to Dan WU, Email: danwoo@126.com, ORCID: 0000-0002-2611-7317