Some simple effective approximations to the 2-Poisson model for probabilistic weighted retrieval

By Stephen E. Robertson, S. Walker. 1994

Subfields: 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 Information Retrieval & Search

Citations in the corpus

A Deep Relevance Matching Model for Ad-hoc Retrieval (2016, background): “For example, on ClueWeb-09-Cat-B topic titles, the relative improvement of our model over the best perform-Table ing baseline (i.e., BM25) is about 11 .”
Learning to Ask: Neural Question Generation for Reading Comprehension (2017, methodology): “, 2016) using three automatic evaluation metrics, we find that our system significantly outperforms a collection of strong baselines, including an information retrieval-based system (Robertson and Walker, 1994), a statistical machine translation approach (Koehn et al.”
Improving Machine Translation Performance by Exploiting Non-Parallel Corpora (2005, neutral): “Utiyama et al. (2003) use the BM25 (Robertson and Walker 1994) similarity measure.”
Probabilistic models of information retrieval based on measuring the divergence from randomness (2002, methodology): “(8) We show in the following sections that the two approximations P and D are experimentally equivalent under the Probability Ranking Principle [Robertson 1986].”
TTPDrill: Automatic and Accurate Extraction of Threat Actions from Unstructured Text of CTI Sources (2017, methodology): “We use an enhanced BM25 weighting function for the TF-IDF method[25] to match candidate actions to threat actions in the”
A cross-benchmark comparison of 87 learning to rank methods (2015, methodology): “Research in the ﬁeld of ranking models has long been based on manually designed ranking functions, such as the well-known BM25 model (Robertson & Walker, 1994).”
Direct Maximization of Average Precision by Hill-Climbing, with a Comparison to a Maximum Entropy Approach (2004, methodology): “For any document, the model predicts the probability of relevance for that document based on the Okapi term frequency (tf ) scores (Robertson and Walker, 1994) for the query terms within it.”
A formal study of information retrieval heuristics (2004, methodology): “The Okapi formula is another highly effective retrieval formula that represents the classical probabilistic retrieval model [8].”
A Survey of Text Clustering Algorithms (2012, background): “One effective way to assign weights to terms when representing a document as a weighted term vector is the BM25 term weighting method [78], where the normalized TF not only addresses length normalization, but also has an upper bound which improves the robustness as it avoids overly rewarding the…”
To Clean or Not to Clean (2018, methodology): “To test the effect of document preprocessing on probabilistic and information theoretic methods, we also experimented with BM25 [60, 61] and Divergence from Randomness [2] retrieval models.”
A dynamic bayesian network click model for web search ranking (2009, unknown): “Labs Yahoo!”
Response Ranking with Deep Matching Networks and External Knowledge in Information-seeking Conversation Systems (2018, methodology): “multi-turn conversation response ranking method as the following: BM25. This method uses the dialog context as the query to retrieve response candidates for response selection. We consider BM25 model [24] as the retrieval model. ARC-II. ARC-II is an interaction focused deep text matching architectur”
Interpreting TF-IDF term weights as making relevance decisions (2008, methodology): “We have tested the IDF factor used in the BM11 term weight and the IDF here.”
SparTerm: Learning Term-based Sparse Representation for Fast Text Retrieval (2020, methodology): “…terms and expand some terms not appearing in the passage but very semantically relevant, even the terms in the target query such as “sign”. as bag-of-words (BoW), such as TF-IDF [15] and BM25 [14], can efficiently perform literal matching, thus playing a core role in industrial IR systems.”
Yahoo! Learning to Rank Challenge Overview (2010, methodology): “Web page ranking has traditionally been based on a manually designed ranking function such as BM25 (Robertson and Walker, 1994).”
An anatomy for neural search engines (2019, methodology): “Handcrafting ranking formulas like BM25 is not an easy task; we are most of times tuning it based on our intuitions and extensive experimentation.”
Learning to Rank for Information Retrieval and Natural Language Processing (2011, unknown): “A>B, B>C) – Implicit relevance judgment: derived from click-through data • Creation of List – List (or permutation) of documents is given – Ideal but difficult to implement Implicit Relevance Judgment Doc A Doc B Doc C B > A ranking of documents at search system users often clicked on Doc B ordered ”
A Dual Embedding Space Model for Document Ranking (2016, background): “In the probabilistic approach, the 2-Poisson model forms the basis for counting term frequency [6, 15, 36].”