Statistical Language Models for Information Retrieval: A Critical Review

By Chengxiang Zhai. Foundations and Trends in Information Retrieval, 2008

453 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Information Systems

View Paper: https://www.semanticscholar.org/paper/6bc5142acd2798eca4b0f8bd96d245f846609bba

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. W. Bruce Croft
2. Jianfeng Gao
3. Hamed Zamani
4. Jimmy J. Lin
5. Hang Li
6. Mohit Bansal
7. Jun Xu
8. Kyunghyun Cho

In Information Systems

1. Hamed Zamani
2. W. Bruce Croft
3. Jun Xu
4. Hang Li
5. Jian-Yun Nie
6. Yoshua Bengio
7. Jimmy J. Lin
8. Debasis Ganguly

Citations in the corpus

• Beyond [CLS] through Ranking by Generation (2020, background): “On the other hand, classical literature on probabilistic models for IR showed that language modeling, a type of simple generative model, can be effective for document ranking (Zhai, 2008; Lafferty and Zhai, 2001; Ponte and Croft, 1998).”
• To Clean or Not to Clean (2018, methodology): “We selected language model–based retrieval methods, with two different smoothing techniques, namely Jelinek-Mercer smoothing, and Dirichlet smoothing [75–77].”
• Learning joint query interpretation and response ranking (2012, methodology): “Given the success of generative techniques in corpus modeling [8], IR [41] and entity ranking [3, 4], it is natural to propose a generative language model approach to joint query interpretation and response ranking.”
• Semantic Matching in Search (2014, methodology): “Traditionally, query document matching is represented in the IR models such as VSM [155], BM25 [153], and LM4IR [204].”
• Pseudo-Relevance Feedback Based on Matrix Factorization (2016, methodology): “In the probabilistic retrieval models, it is assumed that in the non-relevancy conditions, document and query are independent [42], which is similar to our assumption.”
• TILDE: Term Independent Likelihood moDEl for Passage Re-ranking (2021, unknown): “…20] + +++ Deep LMs Re-ranker [18, 19] +++ - - Deep QLMs Re-ranker [8, 33] ++ - - - TILDE (ours) ++ +++ Unlike traditional query likelihood models [29, 30] and deep query likelihood models [8, 33] which only optimise query likelihood during training, BiQDL is used to train TILDE to maximise both…”
• Learning Concept Embeddings for Query Expansion by Quantum Entropy Minimization (2014, methodology): “Query expansion In order to evaluate the effectiveness of the proposed approach and the baselines, we perform QE using the powerful KL-divergence framework (Zhai 2008).”
• PubMed related articles: a probabilistic topic-based model for content similarity (2007, unknown): “Despite the popularity and performance of language modeling techniques for information retrieval (see [14] for an overview), bm25 remains a competitive baseline.”
• Winning Isn't Everything: Training Human-Like Agents for Playtesting and Game AI (2019, methodology): “To build our Markov agent we use a direct approach to style reproduction inspired by natural language processing literature (see review [51]).”
• A Survey of Model Architectures in Information Retrieval (2025, unknown): “The commonly used smoothing methods (Zhai and Lafferty, 2004; Zhai et al., 2008) include Jalinek-Mercer smoothing, Dirichlet smoothing, etc.”
• Measuring the Feasibility of a Question and Answering System for the Sarawak Gazette Using Chatbot Technology (2025, unknown): “These were good enough to enhance IR tasks and general NLP tasks (Liu & Croft, 2005; Zhai, 2008), but these models suffered from the curse of dimensionality as predicting the next word based on previous words required exponentially more computational power and memory (Zhao et al, 2023).”
• Data Kernel Perspective Space Performance Guarantees for Synthetic Data from Transformer Models (2026, unknown): “…models have enabled unprecedented improvement in performance of language technologies for a wide range of tasks (e.g. automatic speech recognition ( [1], [2], [3], [4], [5]), information retrieval ( [6], [7], [8], [9], [10]), particularly since the advent of transformer-based approaches [11], [12].”