Semantic term matching in axiomatic approaches to information retrieval

By Hui Fang, Chengxiang Zhai. 2006

145 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

In Information Systems

In Information Retrieval & Search

Citations in the corpus

A Deep Relevance Matching Model for Ad-hoc Retrieval (2016, background): “For example, Fang and Zhai [7] proposed the semantic term matching constraint which states that matching an original query term exactly should always contribute no less to the relevance score than matching a semantically related term multiple times.”
Simple Techniques for Cross-Collection Relevance Feedback (2019, methodology): “The rows show effectiveness with “bag of words” BM25, BM25 combined with RM3 expansion [1], and BM25 with axiomatic semantic term matching [3]; all used default Anserini parameters.”
Improvements that don't add up: ad-hoc retrieval results since 1998 (2009, background): “Finally, to take an example with mid-range performance, Fang and Zhai [2006] implement their own baselines over which they achieve statistically signiﬁcant improvements.”
ReQ-ReC: high recall retrieval with query pooling and interactive classification (2014, methodology): “Other query reformulation methods can also be applied, such as synonym expansion [24] and semantic term matching [5].”
Semantic Matching by Non-Linear Word Transportation for Information Retrieval (2016, background): “Therefore, a document that can interpret more distinct query words would be assigned a higher score, which is an expected property for IR models [11].”
Dense Text Retrieval Based on Pretrained Language Models: A Survey (2022, background): “Another possible approach to understanding the behavior of dense retrievers is axiomatic analysis from classic IR literature [196]–[198].”
Modeling Diverse Relevance Patterns in Ad-hoc Retrieval (2018, methodology): “uld be able to encode many well-known IR heuristics to be qualified in this layer. According to previous studies, such heuristics at least include the modeling of exact matching and semantic matching [7, 10], proximity [33], term importance [10] and so on. SIGIR ’18, July 8–12, 2018, Ann Arbor, MI, ”