Exercises in computational linguistics . By H. Brandt Corstius. (Mathematical Centre tracts, 30.) Amsterdam: Mathematisch Centrum, 1970. Pp. 162. $5.00.

Joseph E. Grimes · Language · 1971

26 aa 42 h 58 z 11 ay 27 ee 43 k 59 diaeresis(underlining} 12 eeu 28 00 44 b 60 qu 13 el 29 a 45 d 61 ch 14 eui 30 e 46 f 62 apostrophe 15 ey 31 I 47 g 63 case definitions 16 ieu 32 u 48 I 64 hyphenThis code makes it possible to differentiate easily between groups such as "vowels", "syllable delimiters", "condensing letters", and so on.They is considered as a vowel, and in a word like royaal condenses the preceding o to the vowel oy, making a correct hyphenation.The compression matrix is translated into this code as an integer array compression, the empty places get value zero.If one would want to make changes.in the compression (e.g., recognize ea as a vowel, or not combine the digraph ch), then this is possible without changing the program itself.After compression of the oneletters from left to right, the number of syllables is equal to the number of vowels.The only necessary information is therefore: the vowel quality of some generalized letters and the compression matrix.Example: quadraat contains eight oneletters.After compression, it contains six letters: qu, a, d, r, aa and t, of which a and aa are vowels.Therefore, the number of syllables is two.The same procedure is possible for many other languages, with the exception of French where the compression cannot be done in the same wa:y..!!_ h > 4 V h 2 ~ front:= front x 64 + word[word begin + 2]; .!!. h > 3 ~ front:= front x 64 + word[word begin + 3]; f2!. j:= prefixes[h,0] step 1 ~ 18 ~ begin.!!. front = prefixes[h,j] ~ cut off(h) end;

Read the paper · More papers on PaperTik