Ana Luísa Varani Leal

Márcia Schmaltz
Chao Chi Meng
Chiu Lai Fan
Luís Felipe Rodrigues

In Computational Linguistics, work on textual analysis mainly attempts to explain behaviors and relations of linguistic components, properly identified in the textual structure. Such research is especially related to morphological and syntactic structures. Little research has focused on semantic analysis, where the text is the concrete object of study. From textual analysis, the processes related to their constitution are exploited and justified. Textual coherence, which emerges from the rhetorical relations that occur in the textual configuration, is the representative element of the discourse structure. Therefore, it must be analyzed and understood in global terms. In pursuit of this, it is necessary to consider all textual levels related to the process of signification. In order to explore textual coherence, we propose and develop a methodology for the analysis of discourse themes. The proposal was evaluated to identify the localized processes and their relations in the constitution of the subject, aiming to produce an automatic macro-proposition and macro-structure. The results are promising, in terms of precision and recall.
Screenshots of AuTema-Dis I system
Figure 1: Data input.Figure 2: Parser output.

Figure 3: Segmentation.

Figure 4: Rhetorical relations.

Figure 5: Generated macroproposition.

Leal, Ana Luísa Varani, P. Quaresma, Luís Filipe Rodrigues, Maria João Marcalo. 2007. Análise automática de textos em PE e PB – projecto autema-dis: um estudo interdisciplinar em linguística e informática. In XI Congresso Nacional de Lingüística e Filologia. Rio de Janeiro: Cifefil.

Leal, Ana Luísa Varani, Luís Filipe Rodrigues. 2007. Análise Automática de Textos em PE e PB – um estudo interdisciplinar. Comemorações do 75º Aniversário do CLUL 1. 50-56.

Leal, Ana Luísa Varani, P. Quaresma. 2006. DIRPI – um processo interdisciplinar e interinstitucional. In XVI Encontro Da Aulp – Associação das Universidades de Língua Portuguesa, 2006, Macau. Lisboa: AULP.

Leal, Ana Luísa Varani, Paulo Quaresma & Rove Chishman. From syntactical analysis to textual segmentation. In R. Vieira, P. Quaresma, M. G. V. Nunes, N. Mamede, C. Oliveira, & M. C. Dias (eds.) PROPOR, volume 3960 of Lecture Notes in Computer Science, 252–255. Berlin: Springer.

Visual Interactive Syntax Learning
Rhetorical Structure Theory
Daniel Marcu