{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,21]],"date-time":"2026-07-21T07:44:57Z","timestamp":1784619897844,"version":"3.55.0"},"reference-count":19,"publisher":"MIT Press","issue":"10","content-domain":{"domain":["direct.mit.edu"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2022,9,12]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:p>Barlow (1985) hypothesized that the co-occurrence of two events A and B is \u201csuspicious\u201d if P(A,B)\u226bP(A)P(B). We first review classical measures of association for 2 \u00d7 2 contingency tables, including Yule's Y (Yule, 1912), which depends only on the odds ratio \u03bb and is independent of the marginal probabilities of the table. We then discuss the mutual information (MI) and pointwise mutual information (PMI), which depend on the ratio P(A,B)\/P(A)P(B), as measures of association. We show that once the effect of the marginals is removed, MI and PMI behave similarly to Y as functions of \u03bb. The pointwise mutual information is used extensively in some research communities for flagging suspicious coincidences. We discuss the pros and cons of using it in this way, bearing in mind the sensitivity of the PMI to the marginals, with increased scores for sparser events.<\/jats:p>","DOI":"10.1162\/neco_a_01533","type":"journal-article","created":{"date-parts":[[2022,8,26]],"date-time":"2022-08-26T22:05:32Z","timestamp":1661551532000},"page":"2037-2046","update-policy":"https:\/\/doi.org\/10.1162\/mitpressjournals.corrections.policy","source":"Crossref","is-referenced-by-count":7,"title":["On Suspicious Coincidences and Pointwise Mutual Information"],"prefix":"10.1162","volume":"34","author":[{"given":"Christopher K. I.","family":"Williams","sequence":"first","affiliation":[{"name":"School of Informatics, University of Edinburgh, Edinburgh EH8 9AB, U.K. ckiw@inf.ed.ac.uk"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"281","published-online":{"date-parts":[[2022,9,12]]},"reference":[{"key":"2022091222423674700_B1","volume-title":"Categorical data analysis","author":"Agresti","year":"2013"},{"key":"2022091222423674700_B2","doi-asserted-by":"publisher","first-page":"515","DOI":"10.1111\/j.1541-0420.2005.031228.x","article-title":"Frequentist performance of Bayesian confidence intervals for comparing proportions in 2 \u00d7 2 contingency tables","volume":"61","author":"Agresti","year":"2005","journal-title":"Biometrics"},{"key":"2022091222423674700_B3","first-page":"37","article-title":"Cerebral cortex as model builder","volume-title":"Models of the visual cortex","author":"Barlow","year":"1985"},{"key":"2022091222423674700_B4","doi-asserted-by":"crossref","first-page":"395","DOI":"10.1007\/978-94-009-3833-5_18","volume-title":"Matters of Intelligence: Conceptual structures in cognitive neuroscience","author":"Barlow","year":"1987"},{"key":"2022091222423674700_B5","doi-asserted-by":"publisher","first-page":"315","DOI":"10.1007\/s002280050466","article-title":"A Bayesian neural network method for adverse drug reaction signal detection","volume":"54","author":"Bate","year":"1998","journal-title":"European Journal of Clinical Pharmacology"},{"key":"2022091222423674700_B6","article-title":"Normalized (pointwise) mutual information in collocation extraction","volume-title":"Proceedings of the Biennial GSCL Conference 2009","author":"Bouma","year":"2009"},{"issue":"1","key":"2022091222423674700_B7","first-page":"22","article-title":"Word association norms, mutual information, and lexicography","volume":"16","author":"Church","year":"1990","journal-title":"Comput. Linguist."},{"key":"2022091222423674700_B8","author":"Duggirala","year":"2018","journal-title":"Data mining at FDA"},{"issue":"3","key":"2022091222423674700_B9","doi-asserted-by":"crossref","first-page":"177","DOI":"10.1080\/00031305.1999.10474456","article-title":"Bayesian data mining in large frequency tables, with an application to the FDA spontaneous reporting system","volume":"53","author":"DuMouchel","year":"1999","journal-title":"American Statistician"},{"key":"2022091222423674700_B10","volume-title":"Advances in neural information processing systems","author":"Edelman","year":"2002"},{"issue":"1","key":"2022091222423674700_B11","doi-asserted-by":"publisher","first-page":"109","DOI":"10.2307\/2982448","article-title":"The measure of association in a 2 \u00d7 2 table","volume":"126","author":"Edwards","year":"1963","journal-title":"Journal of the Royal Statistical Society, Series A (General)"},{"key":"2022091222423674700_B12","doi-asserted-by":"publisher","first-page":"20","DOI":"10.2174\/1876527001607010020","article-title":"Comparing measures of association in 2 \u00d7 2 probability tables","volume":"7","author":"Hasenclever","year":"2016","journal-title":"Open Statistics and Probability Journal"},{"issue":"1","key":"2022091222423674700_B13","doi-asserted-by":"publisher","first-page":"49","DOI":"10.1093\/genetics\/49.1.49","article-title":"The interaction of selection and linkage. I. General considerations; Heterotic models","volume":"49","author":"Lewontin","year":"1964","journal-title":"Genetics"},{"key":"2022091222423674700_B14","volume-title":"Foundations of statistical natural language processing","author":"Manning","year":"1999"},{"key":"2022091222423674700_B15","volume-title":"Numerical recipes: The art of scientific computing","author":"Press","year":"2007","edition":"3rd"},{"issue":"1","key":"2022091222423674700_B16","doi-asserted-by":"publisher","first-page":"39","DOI":"10.1023\/A:1009713703947","article-title":"Beyond market baskets, Generalizing association rules to dependence rules","volume":"2","author":"Silverstein","year":"1998","journal-title":"Data Mining and Knowledge Discovery"},{"key":"2022091222423674700_B17","doi-asserted-by":"publisher","first-page":"293","DOI":"10.1016\/S0306-4379(03)00072-3","article-title":"Selecting the right objective measure for association analysis","volume":"29","author":"Tan","year":"2004","journal-title":"Information Systems"},{"key":"2022091222423674700_B18","article-title":"On the association of attributes in statistics","volume":"194","author":"Yule","year":"1900","journal-title":"Phil. Trans. Roy. Soc., A"},{"issue":"6","key":"2022091222423674700_B19","doi-asserted-by":"publisher","first-page":"579","DOI":"10.2307\/2340126","article-title":"On the methods of measuring association between two attributes","volume":"75","author":"Yule","year":"1912","journal-title":"Journal of the Royal Statistical Society"}],"container-title":["Neural Computation"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/direct.mit.edu\/neco\/article-pdf\/34\/10\/2037\/2042422\/neco_a_01533.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/direct.mit.edu\/neco\/article-pdf\/34\/10\/2037\/2042422\/neco_a_01533.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,10,2]],"date-time":"2024-10-02T17:40:32Z","timestamp":1727890832000},"score":1,"resource":{"primary":{"URL":"https:\/\/direct.mit.edu\/neco\/article\/34\/10\/2037\/112572\/On-Suspicious-Coincidences-and-Pointwise-Mutual"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,9,12]]},"references-count":19,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2022,9,12]]},"published-print":{"date-parts":[[2022,9,12]]}},"URL":"https:\/\/doi.org\/10.1162\/neco_a_01533","relation":{},"ISSN":["0899-7667","1530-888X"],"issn-type":[{"value":"0899-7667","type":"print"},{"value":"1530-888X","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2022,10]]},"published":{"date-parts":[[2022,9,12]]}}}