Predicting when browsing context is relevant to search

Rahurkar, Mandar; Cucerzan, Silviu

doi:10.1145/1390334.1390532

Cited by 8 publications

(4 citation statements)

References 6 publications

(6 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The core difference from similar classifiers described in previous studies (see e.g. [14,17]) is its direct dependence on the constructed re-ranking algorithm.…”

Section: Query Filteringmentioning

confidence: 95%

Personalization of web-search using short-term browsing context

Ustinovskiy

Serdyukov

2013

Proceedings of the 22nd ACM International Conference on Information &Amp; Knowledge Management

View full text Add to dashboard Cite

Search and browsing activity is known to be a valuable source of information about user's search intent. It is extensively utilized by most of modern search engines to improve ranking by constructing certain ranking features as well as by personalizing search. Personalization aims at two major goals: extraction of stable preferences of a user and specification and disambiguation of the current query. The common way to approach these problems is to extract information from user's search and browsing long-term history and to utilize short-term history to determine the context of a given query. Personalization of the web search for the first queries in new search sessions of new users is more difficult due to the lack of both long-and short-term data.In this paper we study the problem of short-term personalization. To be more precise, we restrict our attention to the set of initial queries of search sessions. These, with the lack of contextual information, are known to be the most challenging for short-term personalization and are not covered by previous studies on the subject. To approach this problem in the absence of the search context, we employ short-term browsing context. We apply a widespread framework for personalization of search results based on the reranking approach and evaluate our methods on the large scale data. The proposed methods are shown to significantly improve non-personalized ranking of one of the major commercial search engines. To the best of our knowledge this is the first study addressing the problem of short-term personalization based on recent browsing history. We find that performance of this re-ranking approach can be reasonably predicted given a query. When we restrict the use of our method to the queries with largest expected gain, the resulting benefit of personalization increases significantly.

show abstract

“…The core difference from similar classifiers described in previous studies (see e.g. [14,17]) is its direct dependence on the constructed re-ranking algorithm.…”

Section: Query Filteringmentioning

confidence: 95%

Personalization of web-search using short-term browsing context

Ustinovskiy

Serdyukov

2013

Proceedings of the 22nd ACM International Conference on Information &Amp; Knowledge Management

View full text Add to dashboard Cite

show abstract

“…There is work on predicting whether a query submitted by a user to a search engine is relevant to a previously browsed news article [37]. Our task is different: predicting which n-grams in a document (e.g., article) are likely to be used as search queries.…”

Section: Related Workmentioning

confidence: 99%

“…The documents in which we recommend queries are news articles from Yahoo! News as they were shown to trigger many search queries [19,37]. Our data includes articlequery pairs where a query was issued to the search engine by at least three users in a day during the first five minutes after reading the article; no other actions were performed in between by the users.…”

Section: Evaluation 41 Experimental Settingmentioning

confidence: 99%

Recommending Search Queries in Documents Using Inter N-Gram Similarities

Sheetrit¹,

Fyodorov²,

Raiber³

et al. 2021

Proceedings of the 2021 ACM SIGIR International Conference on Theory of Information Retrieval

View full text Add to dashboard Cite

Reading a document can often trigger a need for additional information. For example, a reader of a news article might be interested in information about the persons and events mentioned in the article. Accordingly, there is a line of work on recommending search-engine queries given a document read by a user. Often, the recommended queries are selected from a query log independently of each other, and are presented to the user without any context. We address a novel query recommendation task where the recommended queries must be n-grams (sequences of consecutive terms) in the document. Furthermore, inspired by work on using inter-document similarities for document retrieval, we explore the merits of using inter n-gram similarities for query recommendation. Specifically, we use a supervised approach to learn an inter n-gram similarity measure where the goal is that n-grams that are likely to serve as queries will be deemed more similar to each other than to other n-grams. We use the similarity measure in a wide variety of query recommendation approaches which we devise as adaptations of ad hoc document retrieval techniques. Empirical evaluation performed using data gathered from Yahoo!'s search engine logs attests to the effectiveness of the resultant recommendation methods.

show abstract

“…When the users are involved in fact finding and information gathering tasks, they usually submit many short queries, visiting several domains in complex sense-making tasks [92,35]. They typically show various needs at different times based on current circumstances [66,49]. Pages often contain mixtures of topics that are not necessarily interrelated one another, and relevant information is often described over a series of connected blocks which make the cue extraction entangled [93].…”

mentioning

confidence: 99%

Modeling user interests from web browsing activities

Gasparetti

2016

Data Min Knowl Disc

View full text Add to dashboard Cite

Browsing sessions are rich in elements useful to build profiles of user interests, but at the same time HTML pages include noisy data such as advertisements, navigation menus and privacy notes. Moreover, some pages cover several different topics making it difficult to identify the most relevant to the user. For these reasons, they are often ignored by personalized search and recommender systems. We propose a novel approach for recognizing valuable text descriptions of current user information needs—namely cues—based on the data mined from browsing interactions over the web. The approach combines page clustering techniques based on Document Object Model-based representations for acquiring evidence about relevant correlations between text contents. This evidence is exploited for better filtering out irrelevant information and facilitating the construction of interest profiles. A comparative framework proves the accuracy of the extracted cues in the personalize search task, where results are re-ranked according to the last browsed resources

show abstract

Predicting when browsing context is relevant to search

Cited by 8 publications

References 6 publications

Personalization of web-search using short-term browsing context

Personalization of web-search using short-term browsing context

Recommending Search Queries in Documents Using Inter N-Gram Similarities

Modeling user interests from web browsing activities

Contact Info

Product

Resources

About