User interaction is a critical issue affecting system effectiveness and user experience. I divide my research in this direction into two types:

  • observational and explanatory studies that examine why users behave in a particular way and what the influencing factors are;
  • user modeling and interactive techniques for helping users.

I consider the two types of studies as two sides of the same coin—the former often sheds lights on the latter, and the latter is the key to building successful systems.


Human Factors in Interactive Search

Main collaborators: Daqing He (Pitt), James Allan (UMass), Diane Kelly (UTK)

I examine the influence of different factors on users and their search behavior, usually based on laboratory user studies. The purpose is to better understand users and their search behavior, such that we can build better interactive search systems to help users.


Example 1: The Influence of Context on Usefulness Judgments

Multilevel (hierarchical) regression analysis – contextual usefulness judgments (called ESR) as dependent variable (M1) and context-independent usefulness judgments (called Usef.) as dependent variable (M2). Independent variables without † are controls.

Purpose: to examine the difference between contextual and context-independent usefulness judgments.
Method: we conducted a laboratory user study and collected 28 users' contextual and context-independent judgments regarding 736 results.
Findings: our study demonstrates a few advantages of contextual usefulness judgments: it captures users’ real-time state of mind and perceptions; it measures how much useful information the user is able to acquire from a result rather than how much there is in the result; it better reflects users’ needs and criteria of useful results during a session, highlighting novelty as a salient factor. However, we also find that users may not be able to correctly assess the credibility of information during a session, which may reduce the reliability of the collected judgments.

Jiepu Jiang, Daqing He, Diane Kelly, and James Allan. Understanding ephemeral state of relevance. In Proceedings of the Second ACM SIGIR Conference on Human Information Interaction and Retrieval (CHIIR '17), 2017.


Example 2: The Influence of Tasks on Search Behavior

A brief summary of findings regarding the differences of search behavior in four types of search tasks.

Purpose: to examine the difference of search behavior in four types of tasks.
Method: a laboratory user study with 20 users; search tasks differ by target product (Factual or Intellectual) and goal (Specific or Amorphous); eye-tracking.
Findings: click on the figure for a summary.

Jiepu Jiang, Daqing He, and James Allan. Searching, browsing, and clicking in a search session: changes in user behavior by task and over time. In Proceedings of the 37th ACM SIGIR Conference on Research & Development in Information Retrieval (SIGIR '14), 2014.


Intelligent Interactive Search Techniques

Main collaborators: James Allan (UMass), Daqing He (Pitt), Shuguang Han (Pitt)

I build intelligent interactive search systems based on user modeling and user behavior prediction. The purpose is to cope with users' search behavior—promoting beneficial activities and avoiding harmful ones.


Example 3: Reducing Click Errors in Web Search

Experimental results show our approach reduces about 10%–20% of the click and skip errors (measured by DCE@5) with a trade off of 2.1% decline in nDCG@10 (a popular measure of search result relevance).

Search engines provide result summaries to help users quickly identify whether or not it is worthwhile to click on a result and read in detail. However, users may visit useless results or skip useful ones. These actions are usually harmful to the user experience, but few considered this problem in search result ranking. We optimize the relevance of results and user click and skip activities at the same time. Experimental results show our approach reduces about 10%–20% of the click and skip errors with a trade off of 2.1% decline in nDCG@10 (a popular measure of search result relevance).

Jiepu Jiang and James Allan. Reducing click and skip errors in search result ranking. In Proceedings of the 9th ACM International Conference on Web Search and Data Mining (WSDM '16), 2016.


Example 4: Supporting Information Reuse in Cross-Device Web Searches

A predictive search interface for cross-device web search: we predict the most likely to-be-reused queries and results and suggest to users when they search on a new device.

Users usually search on one device and then switch to another. When a user resumes her search on the new device, she often needs to revisit the resource accessed on the previous device. We conducted a laboratory user study with six properly-designed search tasks and 24 participants, resulting in hundreds of search sessions and thousands of queries and clicks. We found that information reuse is very common in cross-device search, and it is particularly important at the beginning of a continued session. We also developed effective techniques for predicting reusing behaviors in the continued sessions, which allows us to build intelligent search interface to support cross-device web search.

Shuguang Han, Jiepu Jiang, Daqing He, and Yu Chi. Understanding and Predicting Information Reuse in Cross-Device Web Searches. Under Review, 2017.


Related Publications:

 

Jiepu Jiang, Daqing He, Diane Kelly, and James Allan. Understanding ephemeral state of relevance. In Proceedings of the Second ACM SIGIR Conference on Human Information Interaction and Retrieval (CHIIR '17), 2017.

Jiepu Jiang and James Allan. Reducing click and skip errors in search result ranking. In Proceedings of the 9th ACM International Conference on Web Search and Data Mining (WSDM '16), 2016. (Acceptance rate 18%)

Jiepu Jiang and Chaoqun Ni. What affects word changes in query reformulation during a task-based search session? In Proceedings of the First ACM SIGIR Conference on Human Information Interaction and Retrieval (CHIIR '16), 2016.

Shuguang Han, Daqing He, Zhen Yue, and Jiepu Jiang. Exploring the contextual support for collaborative information retrieval. In Proceedings of the First ACM SIGIR Conference on Human Information Interaction and Retrieval (CHIIR '16), 2016.

Wei Jeng, Jiepu Jiang, and Daqing He. Users’ perceived difficulties and corresponding reformulation strategies in Google voice search. Journal of Library and Information Studies, 14(1), 2016: 1–14.

Jiepu Jiang, Daqing He, and James Allan. Searching, browsing, and clicking in a search session: changes in user behavior by task and over time. In Proceedings of the 37th ACM SIGIR Conference on Research & Development in Information Retrieval(SIGIR '14), 2014. (Acceptance rate 21%)

Zhen Yue, Shuguang Han, Daqing He, and Jiepu Jiang. Influences on query reformulation in collaborative web search. Computer, 47(3), 2014: 46–53.

Shuguang Han, Daqing He, Jiepu Jiang, and Zhen Yue. Supporting exploratory people search: a study of factor transparency and user control. In Proceedings of the 22nd ACM International Conference on Information & Knowledge Management (CIKM '13), 2013. (Acceptance rate 17%)

Jiepu Jiang, Wei Jeng, and Daqing He. How do users respond to voice input errors? Lexical and phonetic query reformulation in voice search. In Proceedings of the 36th ACM SIGIR Conference on Research & Development in Information Retrieval (SIGIR '13), 2013. (Acceptance rate 20%)

Wei Jeng and Jiepu Jiang. Users’ perceived difficulties and corresponding reformulation strategies in voice search. In Proceedings of the 7th Annual Symposium on Human-Computer Interaction and Information Retrieval (HCIR '13), 2013. (short paper)

Zhen Yue, Jiepu Jiang, Shuguang Han, and Daqing He. Where do the query terms come from? An analysis of query reformulation in collaborative web search. In Proceedings of the 21st ACM International Conference on Information and Knowledge Management(CIKM '12), 2012. (poster)

Shuguang Han, Zhen Yue, Jiepu Jiang, and Wei Jeng. IRIS-IPS: an interactive people search system. In Proceedings of the 6th Annual Symposium on Human-Computer Interaction and Information Retrieval (HCIR '12), 2012. (short paper)

Zhen Yue, Shuguang Han, Jiepu Jiang, and Daqing He. Search tactics as means of examining search processes in collaborative exploratory web search. In Proceedings of the 5th CIKM Ph.D. Workshop on Information and Knowledge, 2012.