Commerce search engines have become popular in recent years, as users increasingly search for (and buy) products on the web. In response to an user query, they surface links to products in their catalog (or index) that match the requirements specified in the query. Often, few or no product in the catalog matches the user query exactly, and the search engine is forced to return a set of products that partially match the query. This paper considers the problem of choosing a set of products in response to an user query, so as to ensure maximum user satisfaction. We call this the result enrichment problem in commerce search. The challenge in result enrichment is two-fold: the search engine needs to estimate the extent to which a user genuinely cares about an attribute that she has specified in a query; then, it must display products in the catalog that match the user requirement on the important attributes, but have a similar but possibly non-identical value on the less important ones. To this end, we propose a technique for measuring the importance of individual attribute values and the similarity between different values of an attribute. A novelty of our approach is that we use entire browse trails, rather than just clickthrough rates, in this estimation algorithm. We develop a model for this problem, design and (theoretically) analyze our algorithm for solving it using browse trails, and support our theoretical findings by showing, via experiments conducted on actual user data, that the algorithm performs well in practice. In the course of developing our algorithm, we offer a solution to another problem that might be of independent interest: we give an algorithm for the annotation of web domains by a set of keywords that represent the contents of the domain. 1.

Recurrent event queries (REQ) constitute a special class of search queries occurring at regular, predictable time intervals. The freshness of documents ranked for such queries is generally of critical importance. REQ forms a significant volume, as much as 6 % of query traffic received by search engines. In this work, we develop an improved REQ classifier that could provide significant improvements in addressing this problem. We analyze REQ queries, and develop novel features from multiple sources, and evaluate them using machine learning techniques. From historical query logs, we develop features utilizing query frequency, click information, and user intent dynamics within a search session. We also develop temporal features by time series analysis from query frequency. Other generated features include word matching with recurrent event seed words and time sensitivity of search result set. We use Naive Bayes, SVM and decision tree based logistic regression model to train REQ classifier. The results on test data show that our models outperformed baseline approach significantly. Experiments on a commercial Web search engine also show significant gains in overall relevance, and thus overall user experience. 1

Activity recognition using mobile phones has great potential in many applications including mobile healthcare. In order to let a person easily know whether he is in strict compliance with the doctor’s exercise prescription and adjust his exercise amount accordingly, we can use a smart-phone based activity reporting system to accurately recognize a range of daily activities and report the duration of each activity. A triaxial accelerometer embedded in the smart phone is used for the classification of several activities, such as staying still, walking, running, and going upstairs and downstairs. The model learnt from a specific person often cannot yield accurate results when used on a different person. To solve the cross-people activity recognition problem, we propose an algorithm known as TransEMDT (Transfer learning EMbedded Decision Tree) that integrates a decision tree and the k-means clustering algorithm for personalized activity-recognition model adaptation. Tested on a real-world data set, the results show that our algorithm outperforms several traditional baseline algorithms.

This dissertation would have not been possible without the persistent guidance and encouragement of my mentors. I owe a big ‘thank you ’ to my advisor, Jamie Callan. It was during the Fall of 2004, when I took two half-semester courses taught by Jamie (Digital Libraries and Text Data Mining), that I discovered the field of Information Retrieval. Seven years later, I am writing my dissertation acknowledgments. Jamie’s advice on both research and life in general has been a great source of insight and support. His feedback on research papers, presentations, lecture slides, proposals, and this dissertation has taught me a great deal about communicating effectively. From Jamie’s example, I have learned valuable lessons on research, teaching, and mentoring. I would like to thank Jaime Carbonell, Yiming Yang, and Fernando Diaz for agreeing to be in my thesis committee. Their feedback was critical in making this dissertation stronger. Fernando Diaz helped shape many of the ideas presented in this dissertation. It was during an internship with Fernando at Yahoo! where I began working on vertical selection. I enjoyed it so much I returned for a second internship a year later. I have been fortunate to have had Fernando as a mentor and collaborator ever since.

This paper introduces query-feature graphs, or QF-graphs. QF-graphs encode associations between high-level descriptions of user goals (articulated as natural language search queries) and the specific features of an interactive system relevant to achieving those goals. For example, a QF-graph for the GIMP software links the query “GIMP black and white” to the commands “desaturate ” and “grayscale. ” We demonstrate how QF-graphs can be constructed using search query logs, search engine results, web page content, and localization data from interactive systems. An analysis of QF-graphs shows that the associations produced by our approach exhibit levels of accuracy that make them eminently usable in a range of real-world applications. Finally, we present three hypothetical user interface mechanisms that illustrate the potential of QF-graphs: search-driven interaction, dynamic tooltips, and app-to-app analogy search. Keywords: Query-feature Graph, Search-driven Interaction