The full text of this item is not available at this time because the student has placed this item under an embargo for a period of time. The Libraries are not authorized to provide a copy of this work during the embargo period, even for Texas A&M users with NetID.
Unified Implicit and Explicit Feedback for Multi-Application User Interest Modeling
MetadataShow full item record
A user often interacts with multiple applications while working on a task. User models can be developed individually at each of the individual applications, but there is no easy way to come up with a more complete user model based on the distributed activity of the user. To address this issue, this research studies the importance of combining various implicit and explicit relevance feedback indicators in a multi-application environment. It allows different applications used for different purposes by the user to contribute user activity and its context to mutually support users with unified relevance feedback. Using the data collected by the web browser, Microsoft Word and Microsoft PowerPoint, Adobe Acrobat Writer and VKB, combinations of implicit relevance feedback with semi-explicit relevance feedback were analyzed and compared with explicit user ratings. Our past research show that multi-application interest models based on implicit feedback theoretically out performed single application interest models based on implicit feedback. Also in practice, a multi-application interest model based on semi-explicit feedback increased user attention to high-value documents. In the current dissertation study, we have incorporated topic modeling to represent interest in user models for textual content and compared similarity measures for improved recall and precision based on the text content. We also learned the relative value of features from content consumption applications and content production applications. Our experimental results show that incorporating implicit feedback in page-level user interest estimation resulted in significant improvements over the baseline models. Furthermore, incorporating semi-explicit content (e.g. annotated text) with the authored text is effective in identifying segment-level relevant content. We have evaluated the effectiveness of the recommendation support from both semi-explicit model (authored/annotated text) and unified model (implicit + semi-explicit) and have found that they are successful in allowing users to locate the content easily because the relevant details are selectively highlighted and recommended documents and passages within documents based on the user’s indicated interest. Our recommendations based on the semi-explicit feedback were viewed the same as those from unified feedback and recommendations based on semi-explicit feedback outperformed those from unified feedback in terms of matching post-task document assessments.
Jayarathna, Ukwatta Kankanamalage Sampath (2016). Unified Implicit and Explicit Feedback for Multi-Application User Interest Modeling. Doctoral dissertation, Texas A & M University. Available electronically from