Many mobile devices today are equipped with diversified sensors that enable the acquisition of rich user context (e.g. GPS location, phone activity) for application utilization. With the growing usage of mobile devices in daily life, the problem of conveniently and promptly searching a piece of content that a user has viewed on his/her device before becomes more and more crucial. This paper aims to propose a context‐based query processing framework called UCQP that supports unstructured queries for content search in a user's access history.
Beyond the keywords related to the content properties, a context query in the framework is specified with freeform phrases that describe high‐level mobile contexts of the user at a previous time when the user viewed the searched content.
Experimental results on a prototype system of the framework illustrate its good accuracy and small response time.
To tolerate the incompleteness and inaccuracy in user query texts caused by fading human memory, the authors develop several semantic query parsers that are tailored for different types of contexts using natural language processing and information retrieval techniques. The authors further propose a similarity model to rank the multiple result contents of a query by comparing context entities specified in the query and historical context values associated with each result.
Xue, W. and Deng, H. (2012), "Unstructured queries based on mobile user context", International Journal of Pervasive Computing and Communications, Vol. 8 No. 4, pp. 368-394. https://doi.org/10.1108/17427371211283038Download as .RIS
Emerald Group Publishing Limited
Copyright © 2012, Emerald Group Publishing Limited