Statistical study on a literary Romanian corpus for the beginning and ending of the words

Author

Mitrea, Adrian ; Vlad, Adriana ; Luca, Adrian

Author_Institution

Fac. of Electron., Telecommun. & Inf. Technol., Politeh. Univ. of Bucharest, Bucharest, Romania

fYear

2012

fDate

21-23 June 2012

Firstpage

81

Lastpage

84

Abstract

The paper attempts to investigate the statistical structure of letters and of letter digrams with which the words begin and end, as well as of trigrams that link two successive words. The investigation is carried out on a printed Romanian literary corpus summing up about 12.5 million words. The impact of the orthography and punctuation marks in the language model assigned to the beginning and to the ending of words is considered.

Keywords

grammars; natural language processing; word processing; language model; letter digrams; letters statistical structure; orthography marks; printed Romanian literary corpus; punctuation marks; successive words; trigrams; Analytical models; Artificial intelligence; Educational institutions; Joining processes; Mathematical model; Pragmatics; Probability; m-grams by which words begin and finish; mathematics of natural language; orthography and punctuation marks;

fLanguage

English

Publisher

ieee

Conference_Titel

Communications (COMM), 2012 9th International Conference on

Conference_Location

Bucharest

Print_ISBN

978-1-4577-0057-6

Type

conf

DOI

10.1109/ICComm.2012.6262546

Filename

6262546