Clickthrough-based latent semantic models for web search

By Jianfeng Gao, Kristina Toutanova, Wen-tau Yih. 2011

98 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Information Systems, Information Retrieval & Search

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Jianfeng Gao
2. Li Deng
3. Zhiyuan Liu
4. Mohit Bansal
5. Xiaodong He
6. Jimmy J. Lin
7. Wen-tau Yih
8. Luke S. Zettlemoyer

In Information Systems

1. Bhaskar Mitra
2. Nick Craswell
3. Xiaodong He
4. Li Deng
5. Jianfeng Gao
6. Hamed Zamani
7. Jun Xu
8. Zhiyuan Liu

In Information Retrieval & Search

1. Nick Craswell
2. Bhaskar Mitra
3. Jianfeng Gao
4. Xiaodong He
5. Wen-tau Yih
6. Jimmy J. Lin
7. Li Deng
8. Hamed Zamani

Citations in the corpus

• Representation Learning Using Multi-Task Deep Neural Networks for Semantic Classification and Information Retrieval (2015, neutral): “402 PLSA(Topic=500) (Hofmann, 1999; Gao et al., 2011) 0.”
• Deep Learning: Methods and Applications (2014, methodology): “In the second line of research, click-through data, which consists of a list of queries and the corresponding clicked documents, is exploited for semantic modeling so as to bridge the language discrepancy between search queries and Web documents in recent studies [120, 124].”
• Learning Continuous Phrase Representations for Translation Modeling (2014, methodology): “, clickthrough data extracted from search logs) for query-document matching (Gao et al. 2011).”
• Multi-Task Retrieval for Knowledge-Intensive Tasks (2021, background): “…tokens explicitly and can potentially map para-phrases of completely different tokens to close vectors, performance of early dense retrieval methods was often inferior to term-matching approaches, except when large labelled data is available (Yih et al., 2011; Gao et al., 2011; Huang et al., 2013).”
• Situational Context for Ranking in Personal Search (2017, background): “There are a number of other approaches which use clickthrough data for web ranking (e.g., [17]).”
• Learning deep structured semantic models for web search using clickthrough data (2013, methodology): “[10] report that both BLTM and DPM outperform significantly the unsupervi sed latent semantic models, including LSA and PLSA, in the doc ument ranking task.”
• Applications of Topic Models (2017, methodology): “Loosely aligned models have been applied in information retrieval for query expansion [Gao et al., 2011, 2012]. While the topic models’ application in information retrieval (Chapter 2) focus on a single language, Gao et al. [2011] assume the queries and Web documents are in different “languages”.”
• How Does NLP Benefit Legal System: A Summary of Legal Artificial Intelligence (2020, methodology): “In addition to these term matching methods, other researchers try to utilize meta-information (Medin, 2000; Gao et al., 2011; Wu et al., 2013) to capture semantic similarity.”
• A Dual Embedding Space Model for Document Ranking (2016, methodology): “ty of relevant documents near the query. ability to answer queries based on memorized clicks. There may even be insufficient behaviour data to learn a click-based embedding [18] or a translation model [10, 19]. In these cases it is crucial to model the relationship between the query and the document ”
• Learning Semantic Representations for the Phrase Translation Model (2013, methodology): “DPM (Row 8) is the Discriminative Projection Model described in [9].”
• A Latent Semantic Model with Convolutional-Pooling Structure for Information Retrieval (2014, methodology): “BLTM (Row 7) is the best performer among different versions of the bilingual topic models described in [15].”
• Incorporating Query Term Independence Assumption for Efficient Retrieval and Ranking using Deep Neural Networks (2019, methodology): “Some neural IR models [Huang et al., 2013, Gao et al., 2011] learn low dimensional dense vector representations of query and document that can be computed independently during inference.”
• Semantic Matching in Search (2014, background): “[71], which gives rise to the query-document pairs in click-through data, where documents are represented by their titles.”
• Learning Discriminative Projections for Text Similarity Measures (2011, background): “In a companion paper, we also demonstrated that various topic models including S2Net can enhance the ranking function (Gao et al., 2011).”