109,299 Hits in 6.7 sec

Did We Get It Right? Predicting Query Performance in E-commerce Search [article]

Rohan Kumar, Mohit Kumar, Neil Shah, Christos Faloutsos
2018 arXiv   pre-print
We demonstrate the feasibility and efficacy of such models in accurately predicting query performance. Our classifier is able to achieve an average AUC of 0.75 on a held-out test set.  ...  We study large-scale user interaction logs from Flipkart's search engine, analyze behavioral patterns and build models to classify queries based on user behavior signals.  ...  Priyank Patel and Mr. Subhadeep Maji for their helpful comments.  ... 
arXiv:1808.00239v1 fatcat:5k5ap2ztfzdqhbimfzw5by5fo4

When video search goes wrong

Christoph Kofler, Linjun Yang, Martha Larson, Tao Mei, Alan Hanjalic, Shipeng Li
2012 Proceedings of the 20th ACM international conference on Multimedia - MM '12  
While insight about a query's performance in general could be obtained using the well-known concept of query performance prediction (QPP), we propose a novel approach for predicting a failure of a video  ...  Our context-aware query failure prediction approach uses a combination of user indicators and engine indicators to predict whether a particular query is likely to fail in the context of a particular search  ...  [4] can be considered to be the work most similar to our own since they use the query, search results and user interaction to predict query performance.  ... 
doi:10.1145/2393347.2393395 dblp:conf/mm/KoflerYLMHL12 fatcat:2jm64oxe5jedtfxdj566t6wlne

An Interaction Mining Approach for Classifying User Intent on the Web

Loredana Caruccio, Vincenzo Deufemia, Giuseppe Polese
2015 Proceedings of the 21st International Conference on Distributed Multimedia Systems  
Predicting the goals of internet users can be extremely useful in e-commerce, online entertainment, and many other internet-based applications.  ...  Beyond these methods, in this paper we propose to mine user interaction activities in order to predict the intent of the user during a navigation session.  ...  The data concerning user interactions during web navigation have been encoded into features, which are used by predictive models to characterize user behaviours.  ... 
doi:10.18293/dms2015-045 dblp:conf/dms/CaruccioDP15 fatcat:wqoywhgtqvdu3ecslefe2mumw4

Deep Sequential Models for Task Satisfaction Prediction

Rishabh Mehrotra, Ahmed Hassan Awadallah, Milad Shokouhi, Emine Yilmaz, Imed Zitouni, Ahmed El Kholy, Madian Khabsa
2017 Proceedings of the 2017 ACM on Conference on Information and Knowledge Management - CIKM '17  
Detecting and understanding implicit signals of user satisfaction are essential for experimentation aimed at predicting searcher satisfaction.  ...  Table 4 : 4 ery level SAT prediction. * and & indicate statistical signi cant (p ≤ 0.05) using paired t-tests compared to the best performing feature based baseline and the best performing sequential  ...  details of user interactions to their distributional vectors which are then used to predict user satisfaction for each query. e architecture of our ConvNet for mapping implicit signals to features is  ... 
doi:10.1145/3132847.3133001 dblp:conf/cikm/MehrotraASYZKK17 fatcat:xpdyde5u3bazbnw5v6zcf6jitu

Selective user interaction

Giridhar Kumaran, James Allan
2007 Proceedings of the sixteenth ACM conference on Conference on information and knowledge management - CIKM '07  
Any opinions, findings and conclusions or recommendations expressed in this material are the author(s) and do not necessarily reflect those of the sponsor.  ...  Acknowledgments This work was supported in part by the Center for Intelligent Information Retrieval, in part by the Defense Advanced Research Projects Agency (DARPA) under contract number HR0011-06-C-0023, and  ...  The goal of that work was to predict in advance if a query will result in acceptable values of precision, and take appropriate steps if the query was predicted to fail (have a low AP).  ... 
doi:10.1145/1321440.1321576 dblp:conf/cikm/KumaranA07 fatcat:h2sv4yxikjerrm73xpapalqvgy

Characterizing and predicting search engine switching behavior

Ryen W. White, Susan T. Dumais
2009 Proceeding of the 18th ACM conference on Information and knowledge management - CIKM '09  
We characterize aspects of switching behavior, and develop and evaluate predictive models of switching behavior using features of the active query, the current session, and user search history.  ...  Our findings provide insight into the decision-making processes of search engine users and demonstrate the relationship between switching and factors such as dissatisfaction with the quality of the results  ...  The best performance is obtained for the session features, followed by query features, and user features.  ... 
doi:10.1145/1645953.1645967 dblp:conf/cikm/WhiteD09 fatcat:oje6tlxb6zhk3kqi5sbz5cx6fi

From "Selena Gomez" to "Marlon Brando"

Iris Miliaraki, Roi Blanco, Mounia Lalmas
2015 Proceedings of the 24th International Conference on World Wide Web - WWW '15  
In this paper, we perform a large-scale analysis into how users interact with the entity results returned by Spark.  ...  Based on this analysis, we develop a set of query and user-based features that reflect the click behavior of users and explore their effectiveness in the context of a prediction task.  ...  Using logistic regression, our results demonstrate that user-based features improve significantly the accuracy for the prediction task compared to using only query-based features.  ... 
doi:10.1145/2736277.2741284 dblp:conf/www/MiliarakiBL15 fatcat:zhpb4m6vb5hwjbuvbwrcfjl2le

Ready to buy or just browsing?

Qi Guo, Eugene Agichtein
2010 Proceeding of the 33rd international ACM SIGIR conference on Research and development in information retrieval - SIGIR '10  
We describe extensive experimental evaluation over both controlled user studies, and logs of interaction data collected from hundreds of real users.  ...  The results show that our method is more effective than the current state-of-the-art techniques, both for detection of searcher goals, and for an important practical application of predicting ad clicks  ...  Acknowledgments The authors thank Microsoft Research and Yahoo! Research for partially supporting this work through faculty research grants.  ... 
doi:10.1145/1835449.1835473 dblp:conf/sigir/GuoA10 fatcat:nb2jhqtq7negpmmf4m3i2qjk7i

Why searchers switch

Qi Guo, Ryen W. White, Yunqiao Zhang, Blake Anderson, Susan T. Dumais
2011 Proceedings of the 34th international ACM SIGIR conference on Research and development in Information - SIGIR '11  
An investigation and evaluation of predicting switching causes with various session-level behavioral features.  ...  Using this feedback, we investigate in detail the reasons that users switch engines within a session.  ...  [10] used interaction features, including switching features, to predict query performance. Leskovec et al.  ... 
doi:10.1145/2009916.2009964 dblp:conf/sigir/GuoWZAD11 fatcat:jkybyfnyqzbynmk5ygcfwnkm7m

A Proposal for User-Focused Evaluation and Prediction of Information Seeking Process

Chirag Shah
2013 European Workshop on Human-Computer Interaction and Information Retrieval  
Our prediction method uses a collection of features extracted solely from the search process such as dwell time, query entropy and relevance judgment in order to evaluate whether it will lead to low or  ...  One of the ways IR systems help searchers is by predicting or assuming what could be useful for their information needs based on analyzing information objects (documents, queries) and finding other related  ...  The author is also grateful to his PhD students Chathra Hendahewa and Roberto Gonzalez-Ibanez for their valuable contributions to this work.  ... 
dblp:conf/eurohcir/Shah13 fatcat:vhohzuhliranrn67s4vh3gnqbu

Predicting Search Task Difficulty [chapter]

Jaime Arguello
2014 Lecture Notes in Computer Science  
In addition to user-interaction features used in prior work, we evaluate features generated from scroll and mouse-movement events on the SERP.  ...  and prior knowledge features did not improve performance.  ...  As expected, using all user-interaction features (all), whole-session prediction was more effective than first-round prediction (p < .05).  ... 
doi:10.1007/978-3-319-06028-6_8 fatcat:yyqnm47dxfbu3lhzxxapdbu2mi

To personalize or not to personalize

Jaime Teevan, Susan T. Dumais, Daniel J. Liebling
2008 Proceedings of the 31st annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '08  
We characterize queries using a variety of features of the query, the results returned for the query, and people's interaction history with the query.  ...  Using these features we build predictive models to identify queries that can benefit from personalization.  ...  Queries are characterized using a variety of features of the query, the results returned for the query, and the query"s interaction history.  ... 
doi:10.1145/1390334.1390364 dblp:conf/sigir/TeevanDL08 fatcat:rabybdw6lfbepovqcynyw2bhdu

Does That Mean You're Happy?

Kyle Williams, Imed Zitouni
2017 Proceedings of the 2017 ACM on Conference on Information and Knowledge Management - CIKM '17  
We use a Long Short-Term Memory (LSTM) Recurrent Neural Network (RNN) to model the sequence of user interactions and show that it performs signi cantly be er than other baselines when detecting good abandonment  ...  Examples of these features include answers on the SERP and detailed Web result snippets.  ...  For instance, query reformulation has been used to measure search success [17] and it was shown that predicting satisfaction and success based on query features results in be er performance than predicting  ... 
doi:10.1145/3132847.3133035 dblp:conf/cikm/WilliamsZ17 fatcat:vqwvctxcr5gjzcthqd4mpy5jxu

Detecting Parkinson's Disease from interactions with a search engine: Is expert knowledge sufficient? [article]

Liron Allerhand, Brit Youngmann, Elad Yom-Tov, David Arkadir
2018 arXiv   pre-print
Our results indicate that mouse tracking data can help in detecting users at early stages of the disease, and that both expert-generated features and unsupervised techniques for feature generation are  ...  Here we show that mouse tracking data collected during people's interaction with a search engine can be used to distinguish PD patients from similar, non-diseased users and present a methodology developed  ...  Performance as a function of data availability Here we focus on the model achieving the best prediction results: the one that uses all features, followed by an aggregation-per-user.  ... 
arXiv:1805.01138v1 fatcat:gjcnxfznjnfo5pez2h5b4foysi

Characterizing and Predicting Voice Query Reformulation

Ahmed Hassan Awadallah, Ranjitha Gurunath Kulkarni, Umut Ozertem, Rosie Jones
2015 Proceedings of the 24th ACM International on Conference on Information and Knowledge Management - CIKM '15  
We then train classifiers to distinguish between reformulation and non-reformulation query pairs and to predict the rationale behind reformulation.  ...  We use this data to compare and contrast characteristics of reformulation and non-reformulation queries over a large a number of dimensions.  ...  We also notice that using semantic similarity features results in a performance that surpasses lexical and phrase similarity features (2 nd row).  ... 
doi:10.1145/2806416.2806491 dblp:conf/cikm/AwadallahKOJ15 fatcat:clxhk66gpjcy5n6uccdyw2gbzy
« Previous Showing results 1 — 15 out of 109,299 results