Improving web search ranking by incorporating user behavior information

@article{Agichtein2006ImprovingWS,
  title={Improving web search ranking by incorporating user behavior information},
  author={Eugene Agichtein and E. Brill and S. Dumais},
  journal={Proceedings of the 29th annual international ACM SIGIR conference on Research and development in information retrieval},
  year={2006}
}
  • Eugene Agichtein, E. Brill, S. Dumais
  • Published 2006
  • Computer Science
  • Proceedings of the 29th annual international ACM SIGIR conference on Research and development in information retrieval
We show that incorporating user behavior data can significantly improve ordering of top results in real web search setting. We examine alternatives for incorporating feedback into the ranking process and explore the contributions of user feedback compared to other common web search features. We report results of a large scale evaluation over 3,000 queries and 12 million user interactions with a popular web search engine. We show that incorporating implicit feedback can augment other features… Expand
Learning user interaction models for predicting web search result preferences
TLDR
This work presents a real-world study of modeling the behavior of web search users to predict web search result preferences and generalizes the approach to model user behavior beyond clickthrough, which results in higher preference prediction accuracy than models based on clickthrough information alone. Expand
Analysis of Desktop Search and Ranking of Their Results Based on Semantics from User Feedback
  • Mehul Smriti Raje
  • Computer Science
  • 2016 12th International Conference on Signal-Image Technology & Internet-Based Systems (SITIS)
  • 2016
TLDR
A Bayesian technique to collect user feedback and use it to rank the results of existing search frameworks is discussed, showing that incorporating user feedback improves the ranking of results by as much as 7.32%. Expand
A Personalized URL Re-ranking Methodology Using User's Browsing Behavior
TLDR
A weight computation method that assigns different values to anchor texts according to user's browsing behaviors such as 'clicking' or 'downloading' is proposed and can be practical for saving surfing time and effort to find users' preferred web pages. Expand
Evaluating the accuracy of implicit feedback from clicks and query reformulations in Web search
TLDR
It is found that relative preferences derived from clicks are reasonably accurate on average, and not only between results from an individual query, but across multiple sets of results within chains of query reformulations. Expand
Aggregated click-through data in a homogeneous user community
TLDR
This study examines the search behaviour of users in a close-knit community on common queries to argue that the benefit of using aggregated clickthrough data varies from task to task: it may improve document rankings for navigational or specific informational queries, but is less likely to be of value to users issuing a broad informational query. Expand
Adaptive Semantic Search: Re-Ranking of Search Results Based on Webpage Feature Extraction and Implicitly Learned Knowledge of User Interests
TLDR
This paper proposes a method of re-ranking the search results that have been primarily ranked using either conventional algorithms that use link structure and user clicks or semantic algorithms, using a combination of general webpage features and user interests. Expand
An Efficient Hybrid Usage-based Ranking model for information retrieval systems & web search Engine
TLDR
An Efficient Hybrid Usage-based Ranking Algorithm called EHURA was applied to 1033 English Corpus to measure its performance and the result shows improvement of the precision over the Content-based ranking algorithm representation while realizing approximately the same recall percentage. Expand
Entity ranking using click-log information
TLDR
This work presents a novel framework for feature extraction that is based on the notions of entity matching and attribute frequencies and introduces different methods and metrics for ranking, which combine them with existing traditional techniques and are studied using real and synthetic data. Expand
Time spent on a web page is sufficient to infer a user's interest
TLDR
Experiments indicate that the most important criterion to gauge the user's interest is the time he or she has spent in a given web site in the past. Expand
A Case Study of User Behavior in Information Retrieval
TLDR
A case study of various approaches in using user behavior for information retrieval is provided and it is suggested that taking absolute feedback about whether the retrieved documents are relevant or not is very difficult, and if the authors can take the implicit feedback from the user in the form of user behavior, they can arrive at a better learning function for the algorithm. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 32 REFERENCES
Learning user interaction models for predicting web search result preferences
TLDR
This work presents a real-world study of modeling the behavior of web search users to predict web search result preferences and generalizes the approach to model user behavior beyond clickthrough, which results in higher preference prediction accuracy than models based on clickthrough information alone. Expand
Query chains: learning to rank from implicit feedback
TLDR
A novel approach for using clickthrough data to learn ranked retrieval functions for web search results by using query chains to generate new types of preference judgments from search engine logs, thus taking advantage of user intelligence in reformulating queries. Expand
Optimizing web search using web click-through data
TLDR
A novel iterative reinforced algorithm to utilize the user click-through data to improve search performance and effectively finds "virtual queries" for web pages and overcomes the challenges discussed above. Expand
Implicit feedback for inferring user preference: a bibliography
TLDR
Traditional relevance feedback methods require that users explicitly give feedback by specifying keywords, selecting and marking documents, or answering questions about their interests, which can be difficult to collect the necessary data and the effectiveness of explicit techniques can be limited. Expand
Evaluating the Robustness of Learning from Implicit Feedback
TLDR
A model of user behavior is created by drawing upon user studies in laboratory and real-world settings and it is found that learning from implicit feedback can be surprisingly robust. Expand
Optimizing search engines using clickthrough data
TLDR
The goal of this paper is to develop a method that utilizes clickthrough data for training, namely the query-log of the search engine in connection with the log of links the users clicked on in the presented ranking. Expand
Evaluating implicit measures to improve web search
TLDR
There was an association between implicit measures of user activity and the user's explicit satisfaction ratings, and the best models for individual pages combined clickthrough, time spent on the search result page, and how a user exited a result or ended a search session. Expand
Implicit Feedback for Recommender Systems
TLDR
This paper identifies three types of implicit feedback and suggests two strategies for using implicit feedback to make recommendations. Expand
IR evaluation methods for retrieving highly relevant documents
TLDR
The novel evaluation methods and the case demonstrate that non-dichotomous relevance assessments are applicable in IR experiments, may reveal interesting phenomena, and allow harder testing of IR methods. Expand
HARD Track Overview in TREC 2003: High Accuracy Retrieval from Documents
Abstract : The High Accuracy Retrieval from Documents (HARD) track explores methods for improving the accuracy of document retrieval systems. It does so by considering three questions. Can additionalExpand
...
1
2
3
4
...