Computational linguistics
Atiyeh Kamyabi Gol; Elham Akhlaghi; Hanieh Habibi
Abstract
A corpus consists of natural continuous texts that are presented electronically and from which information about linguistic elements (both lexical and non-lexical) can be extracted; this can be done in the shortest possible time and with the highest degree of accuracy. The authors of this article have ...
Read More
A corpus consists of natural continuous texts that are presented electronically and from which information about linguistic elements (both lexical and non-lexical) can be extracted; this can be done in the shortest possible time and with the highest degree of accuracy. The authors of this article have created Ferdowsi Annotated Academic Language Corpus and the purpose of this article is to introduce the corpus of faculty members of Ferdowsi University of Mashhad and use it in analyzing and describing hedges in both fields of humanities and science. Hedges indicate the certainty of the author through words such as "maybe", "possibly" and "definitely". Research in this area has mainly focused on manual analysis of limited number of words. In this study, Salager Meyer’s (1997) classification has been used to identify, and categorize hedges. The data were extracted from Ferdowsi Annotated Academic Language Corpus which contains 1100 Persian language articles. Based on the results, the use of hedges in the field of humanities was about twice as much as in the field of science. Meanwhile, the use of conditionals and expressions of doubt in both fields had a high frequency. The results from this study have a better chance at being generalized to the fields under study due to the greater number of entries. The accuracy of the corpus labeling is 96% and it can be used as an excellent source for analysis and research on scientific texts.
