Meu SciELO
Serviços Personalizados
Artigo
Indicadores
- Citado por SciELO
Links relacionados
- Similares em SciELO
Compartilhar
Revista Cubana de Ciencias Informáticas
versão On-line ISSN 2227-1899
Resumo
GONZALEZ, Camila et al. Discovery of knowledge in the comments made by users in a digital news system. Rev cuba cienc informat [online]. 2019, vol.13, n.4, pp. 56-66. ISSN 2227-1899.
Opinion mining is a process of extracting new knowledge and unstructured textual data using automatic methods of detecting and extracting opinions. It has shown a great boom in recent years, due to the need to analyze a large number of opinions and comments that are published daily. The main problem with the data collected from a web news system is that they are in unstructured form, which makes it difficult to identify public opinion and, in turn, the feeling it conveys. The objective of this research focused on discovering knowledge to determine the intentionality of users through their comments on articles on different topics of society. In order to carry out data processing and data transformation, a dataset was initially created with the comments of 5 articles selected by higher level of comments. This data source was used as training for automatic learning. For this purpose, after a study on the main algorithms of natural language processing and opinion mining for the analysis of feelings, a vectorial support machine was specifically chosen. It was obtained as a result, by means of the classification methods, if the users had positive, negative or neutral valuations with respect to the published articles.
Palavras-chave : sentiment analysis; discovery of knowledge; opinion mining; natural language processing.