Building a Corpus for Personality-dependent Natural Language Understanding and Generation

Building a Corpus for Personality-dependent Natural Language Understanding and Generation R.M.S. Ramos, G.B.S. Neto, B.B.C. Silva, D.S. Monteiro, I. Paraboni, R.F.S. Dias University of S˜ao Paulo, School of Arts, Sciences and Humanities S˜ao Paulo, Brazil {ricellimsilva,georges.stavracas,barbarab.claudino,danisamon,ivandre.paraboni,rafaelsandronidias}@gmail.com Abstract The computational treatment of human personality - both for the recognition of personality traits from text and for the generation of text so as to reflect a particular set of traits - is central to the development of NLP applications. As a means to provide a basic resource for studies of this kind, this article describes the b5 corpus, a collection of controlled and free (non-topic specific) texts produced in different (e.g., referential or descriptive) communicative tasks, and accompanied by inventories of personality of their authors and additional demographics. The present discussion is mainly focused on the various corpus components and on the data collection task itself, but preliminary results of personality recognition from text are presented in order to illustrate how the corpus data may be reused. The b5 corpus aims to provide support for a wide range of NLP studies based on personality information and it is, to the best of our knowledge, the largest resource of this kind to be made available for research purposes in the Brazilian Portuguese language. Keywords: Corpora, Personality, Big Five 1. Introduction In recent years, the development of so-called intelligent systems has devoted a great deal of attention to the compu- tational treatment of human personality. This interest may be explained, among other reasons, by the practical obser- vation that users of computer systems not only attribute hu- man traits to the systems they interact with, but they also prefer those systems that present traits similar to their own (Mairesse et al., 2007). Fundamental personality traits are consistently reflected in the language choices made by individuals when communi- cating. For instance, an individual with narcissistic traits might make frequent use of first-person expressions (‘I’, ‘for me’, etc.). The relation between personality and nat- ural language is the focus of a large body of work in the Psychology field, and it is perhaps best summarised by the Big Five personality factors (Goldberg, 1990) - Openness to experience, Conscientiousness, Extraversion, Agreeable- ness and Neuroticism - which are widely accepted as an adequate basis for the representation of human personality. Given its linguistic motivation, the Big Five model provides a theoretical basis for the computational treatment of per- sonality on at least two fronts: the automatic recognition of personality traits from text (which is a language under- standing task), and the generation of text in order to repro- duce certain personality traits of interest (which is a natural language generation (NLG) task). Knowing the personality traits of an individual (e.g., from his/her social network sta- tus updates) has many obvious applications, including staff recruitment, credit analysis etc. In addition to that, person- ality information may also guide the automatic generation of personalised content, the modelling of psychologically plausible virtual agents (e.g., intelligent tutors, game char- acters, etc.) and …

Content truncated to 500 words