Publicação do professor Tiago de Melo.
Resumo (abstract) [1]:
A large number of opinions on products and their features are posted every day on e-commerce websites in user reviews. They are a valuable source of knowledge for both manufacturers and customers. However, reviews often bring so much information that exceeds the human capacity of reasoning and hampers their effective use. Thus, researchers on how to organize a large number of opinions available on the reviews in the Web play a substantial role. Traditional summarization methods group opinions around aspects, but they tend to generate too many aspects groups that are generic and difficult to interpret. We claim that the most important characteristics of the products correspond to the attributes in product catalogs. Thus, these attributes should guide the process of organizing opinions. This paper presents a summary of an approach called OpinionLink, based on machine-learning techniques, to enrich a product catalog with opinions extracted from product reviews. The experimental results demonstrate the effectiveness of the proposed approach.
[1] https://link.springer.com/chapter/10.1007%2F978-3-030-52485-2_17
0 comentários