Improving Mobile Web Search by Clustering and Visualizing Search Engine Results

The search engine makes vast amounts of information available to mobile web users. Despite users' increasing reliance on mobile search engines to meet their information needs, the mobile web search remains challenging due to the small screen of handheld devices and the intense scrolling inherent in the current presentation of search engine results (SERs). The presentation of mobile SERs is important to facilitate their effective search and browsing. The primary objective of this study is to improve the presentation of mobile SERs by grouping them into meaningful clusters and visualizing the clusters.;Based on the cognitive load theory and information foraging theory, we proposed hypotheses about the possible effects of visualization and synchronous presentation of clusters and SERs on user search performance and perception. Moreover, based on categorization of search tasks and the task-technology fit theory, we examined the possible moderating effects of task type on the effects of visualization and synchronicity of SERs presentation. To test the hypotheses, we implemented five mobile web search systems. The proposed mobile web search systems were evaluated using a mixture of quantitative and qualitative methods.;The results of the experiment show that neither the visualization nor the synchronicity of presentation of clusters and SERs had a main effect on the performance of mobile web search. However, when the open-ended and close-ended search tasks were examined separately, the combination of visualization and synchronous presentation of clusters and SERs was found to improve the F-measure and some satisfaction items for the open-ended search tasks.