QUEMDISSE? Reported speech in Portuguese

QUEMDISSE? Reported speech in Portuguese Cl´audia Freitas, Bianca Freitas, Diana Santos PUC-Rio & Linguateca, PUC-Rio, Linguateca & University of Oslo claudiafreitas@puc-rio.br, bianca.frejes@gmail.com, d.s.m.santos@ilos.uio.no Abstract This paper presents some work on direct and indirect speech in Portuguese using corpus-based methods: we report on a study whose aim was to identify (i) Portuguese verbs used to introduce reported speech and (ii) syntactic patterns used to convey reported speech, in order to enhance the performance of a quotation extraction system, dubbed QUEMDISSE?. In addition, (iii) we present a Portuguese corpus annotated with reported speech, using the lexicon and rules provided by (i) and (ii), and discuss the process of their annotation and what was learned. Keywords: quotation verbs; reported speech; corpus annotation; Portuguese 1. Introduction and Motivation A considerable amount of language activities involves re- porting what others have said. In certain contexts, such as the journalistic discourse, the use of reported speech is cru- cial. (Bergler et al., 2004) found that there are pieces of news in which over 90% of the sentences include a quota- tion. In natural language processing (NLP), automatic identifica- tion of reported speech is called Quotation Extraction (QE), which aims to identify quotations in text, and relate them to their authors. This is a task often associated, or subsidiary, to sentiment analysis, and it is the distinctive subtask for “opinion-oriented information extraction” (Pang and Lee, 2008). Its focus is to identify what is said, who said it, and the speaker’s and the writer’s judgments on what was said. New contexts for the exploration of reported speech have recently been provided by research in Digital Humanities, see (Mambrini et al., 2012). Through reported speech, one gives voice to different characters both in a piece of news or in fiction. Consequently, both the presence and the ab- sence of particular characters indicate choices by the text author. By means of quotation identification, it is possi- ble to measure which characters are given voice and, by contrast, which ones are silenced. (Smith et al., 2014), re- searching female characters in popular movies in 11 coun- tries, established, among other findings, that only 23% of these women have lines in action movies. Exploring large amounts of texts – either news or fictional works – and who they quote or silence may provide us with new find- ings about our society. Given a relatively regular structure of reported speech – although quite different in different languages, and even across varieties, see (Santos, 1998) for some discussion –, rule-based approaches to QE are often extremely suc- cessful. However, purely formal marks that indicate the presence of a quotation, such as quotes in English, are not unique to this purpose, hence recognizing the specific verbs that are used in these contexts is highly relevant. Addition- ally, not all reported speech has the aforementioned formal marks. Indirect quotations, constituting almost half of the reported speech in English news (Pareti et al., 2013), are more difficult to identify, and not always covered by QE systems. On …

Content truncated to 500 words