An Analysis of English Punctuation

Abstract

Punctuation has usually been ignored by researchers in computational linguistics over the years. Recently, it has been realized that a true understanding of written language will be impossible if punctuation marks are not taken into account. This paper contains the details of a computer-aided exercise to investigate English punctuation practice for the special case of comma (the most significant punctuation mark) in a parsed corpus. The study classifies the various "structural" uses of the comma according to the syntax-patterns in which a comma occurs. The corpus (Penn Treebank) consists of syntactically annotated sentences with no part-of-speech tag information about the individual words.

An Information-Based Treatment of Punctuation in Discourse Representation Theory

“…The reader is referred to (Say 1995) for a detailed review of punctuation. Other recent papers of our group which may also be useful include (Bayraktar et al 1996;Say and Akman 1996b;Say and Akman 1996a) .…”

Current approaches to punctuation in computational linguistics

“…For tuning the punctuation modules in NLP, more metrics of the sort punctuation-complexity (Jones, 1994b) and stability (Bayraktar, 1996) would be useful in characterizing specific kinds of texts.…”

“…Punctuation for the rhetorical and logical structure of the text became so widespread that 19th and early 20th century novelists frequently used punctuation as one of the features of the written medium to create illusions and "stream of consciousness" effects (Parkes, 1993, p. 87). As can be seen in the works emphasising the usage of punctuation marks in modern texts (Bayraktar, 1996;Jones, 1995;Meyer, 1986), punctuation is still an integral part of the written language. A typical English sentence is likely to contain three or four punctuation symbols, and a punctuation mark of some variety is likely to be encountered on average every fourth to seventh word (Jones, 1997, p. 87).…”

“…He made a percentage analysis of various punctuation marks used as well as comparing complexity and genre of the texts with the frequency of the marks. Bayraktar (1996) has conducted another computational study on the usage of comma, taking classes (categories of use) as specified by Ehrlich (1992) as a basis for matchiag 241 syntax-patterns of comma uses in the Wall Street Journal Corpus (ACL/DCI, 1991) against them. By devising a metric called stability, he shows how the standard usage of a semantic class can be assessed.…”

A more precise analysis of punctuation for broad-coverage surface realization with CCG

“…1 Here, the effect of combining the comma with the phrase headed by despite is to add the EMPH-INTRO + feature to the despite-phrase's semantics. Following (Bayraktar et al, 1998), this feature indicates that the comma has the discourse function of emphasizing an introductory clause or phrase. During realization, the feature triggers the look-up of the category in (2), and prevents the re-application of the category to its own output (as the feature should only be realized once).…”