Supporting Informational Search with Click Tracking and Data Visualization Techniques

  • 張 瑋豪

Student thesis: Master's Thesis


In the era of information explosion people commonly use web search to obtain required information When conducting informational web search a user may browse search results returned by the search engine locate the relevant information in need click on interesting entries and scrutinize them Additionally he or she may start a new search session to find more information This process iterates until he or she is satisfied with all obtained information For example when conducting a web search on “keeping a pet ” some people are looking for small-sized pets some others are looking for the most exotic pets and some more others just want to know the responsibility of keeping a pet It is thus observed that there are different search activities for different people even if they start with identical keywords In this work we propose to track the click behavior of a user so as to identify an informational search Note that click behavior is an implicit type of user feedback Moreover we also exploit data visualization techniques and reuse click data to support the process of an informational web search Note that data visualization makes it easier to understand complicated data by illustrations Even if a user does not have a clear search target or appropriate keywords our approach still works by showing relevant keywords to help him or her make decisions Finally volunteers are asked to evaluate the performance of our prototype system by completing several quizzes In each quiz the corresponding time consumption is compared with that of directly using a search engine Experimental results show that our approach help users to accomplish their informational web search in a more efficient way
Date of Award2016 Aug 27
Original languageEnglish
SupervisorWei-Guang Teng (Supervisor)

Cite this