Some Experiments on Clustering Similar Sentences of Texts in Portuguese

Identifying similar text passages plays an important role in many applications in NLP, such as paraphrase generation, automatic summarization, etc. This paper presents some experiments on detecting and clustering similar sentences of texts in Brazilian Portuguese. We propose an evalution framework based on an incremental and unsupervised clustering method which is combined with statistical similarity metrics to measure the semantic distance between sentences. Experiments show that this method is robust even to treat small data sets. It has achieved 86% and 93% of F-measure and Purity, respectively, and 0.037 of Entropy for the best case.