Han, Shuguang
(2018)
UNDERSTANDING, MODELING AND SUPPORTING CROSS-DEVICE WEB SEARCH.
Doctoral Dissertation, University of Pittsburgh.
(Unpublished)
Abstract
Recent studies have witnessed an increasing popularity of cross-device web search, in which users resume their previously-started search tasks from one device to later sessions on another. This novel search mode brings new user behaviors such as cross-device information transfer; however, they are rarely studied in recent research. Existing studies on this topic mainly focused on automatic cross-device search task extraction and/or task continuation prediction; whereas it lacks sufficient understanding of user behaviors and ways of supporting cross-device search tasks. Building an automated search support system requires proper models that can quantify user behaviors in the whole cross-device search process. This motivates me to focus on understanding, modeling and supporting cross-device search processes in this dissertation.
To understand the cross-device search process, I examine the main cross-device search topics, the major triggers, the information transfer approaches, and users’ behavioral patterns within each device and across multiple devices. These are obtained through an on-line survey and a lab-controlled user study with fine-grained user behavior logs. Then, I work on two quantitative models to automatically capture users' behavioral patterns. Both models assume that user behaviors are driven by hidden factors, and the identified behavioral patterns are either the hidden factors or a reflection of hidden factors. Following prior studies, I consider two types of hidden factors --- search tactic (e.g., the tactic of information re-finding/finding would drive to click/skip previously-accessed documents) and user knowledge (e.g., knowing the knowledge within a document would drive users to skip the document). Finally, to create a real-world cross-device search support use case, I design two supporting functions: one to assist information re-finding and the other to support information finding. The effectiveness of different support functions are further examined through both off-line and on-line experiments.
The dissertation has several contributions. First, this is the first comprehensive investigation of cross-device web search behaviors. Second, two novel computational models are proposed to automatically quantify cross-device search processes, which are rarely studied in existing researches. Third, I identify two important cross-device search support tasks and implement effective algorithms to support both of them, which can beneficial future studies for this topic.
Share
Citation/Export: |
|
Social Networking: |
|
Details
Item Type: |
University of Pittsburgh ETD
|
Status: |
Unpublished |
Creators/Authors: |
|
ETD Committee: |
|
Date: |
24 January 2018 |
Date Type: |
Publication |
Defense Date: |
14 April 2017 |
Approval Date: |
24 January 2018 |
Submission Date: |
5 December 2017 |
Access Restriction: |
No restriction; Release the ETD for access worldwide immediately. |
Number of Pages: |
180 |
Institution: |
University of Pittsburgh |
Schools and Programs: |
School of Information Sciences > Information Science |
Degree: |
PhD - Doctor of Philosophy |
Thesis Type: |
Doctoral Dissertation |
Refereed: |
Yes |
Uncontrolled Keywords: |
Cross-device Web Search
Cross-Device Search Process
Cross-device Search Support
Search Process Modeling
Information Re-finding
Information Exploration
Search Tactics
Search as Knowledge Learning |
Date Deposited: |
24 Jan 2018 16:27 |
Last Modified: |
24 Jan 2018 16:27 |
URI: |
http://d-scholarship.pitt.edu/id/eprint/33717 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
|
View Item |