Semiautomatic disabbreviation of technical text

Normalization of non-standard words

“…We have already noted the similarity of the current work to that of Rowe and Laitinen (1995), discussed in Section 2.3. There is also a similarity between the current work and work reported in Taghva and Gilbreth (1995) on using approximate string matching methods to induce interpretations of acronyms or letter sequences from their full word expansions found somewhere in the immediate context of the given letter sequence.…”

“…In particular, if there is a text that only refers to the Dow Jones with the letter sequence DJIA, that text would not be retrieved. Rowe and Laitinen (1995) describe a semiautomatic procedure for guessing the expansion of novel abbreviations in text. Their method depends upon dictionaries of known full words, and dictionaries of known abbreviations and their expansions.…”

Automatic caption localization for photographs on World Wide Web pages

“…Comparisons for F5 and F6 are done after conversion to lower case, and F6 ignores the common words of English not nouns, verbs, adjectives, or adverbs (numbers and 154 special words including "and", "the", "ours", "of", "without", "when"), with exceptions for spatial and temporal prepositions. F6 also checks for abbreviations within the file name of words or pairs of words in the caption with methods from Rowe and Laitinen (1995). http://faculty.nps.edu/ncrowe/marie/webpics.html Figure 3 includes statistics on caption processing, and the lower-left curve in Figure 4 is the recall-precision curve for the output of the caption neuron for the test set after training on the training set.…”

Understanding of Navy Technical Language via Statistical Parsing

“…We have several hundred rules for special formats including multiword ones, which interpret "bu# 462945" as an aircraft identification number by its front word, "sawtooth mountains" as mountains by its tail word, "02/21/93" as a date, "10-20m" as a range of meters, "visit/dedication" as a conjunction, and "shiploading" as a noun-gerund equivalent of an adjective. Misspellings and abbreviations were obtained mostly automatically, with human post-checking, using the methods described in (Rowe and Laitinen, 1995). Lexical ellipsis (e.g.…”