Identifying Social and Expressive Factors in Request Texts Using Transaction/Sequence Model

The paper focuses on requests in written forms, where we describe a novel approach to computational modelling of specific features of politeness in speech act of requesting. We examine the similarities and differences in the use of specific social and expressive factors in two languages (mother tongue and a foreign language). The requests collected from different social situations among students and their teachers in a university environment were used as data source for a research. Transaction/Sequence model for text representation was formulated and association rules analysis was applied as a research method. The findings are interesting mainly in terms of differences in the use of politeness features in foreign language and mother tongue. The results indicated that the requests written in mother tongue are less direct than in foreign language.