Abstract

ILLANES PEREDO, Daniel Alejandro; ARTEAGA SABJA, Wendoline  and  SANDOVAL ALCOCER, Juan Pablo. Improving contact information search on Facebook analyzing emotions. RevActaNova. [online]. 2019, vol.9, n.2, pp.257-270. ISSN 1683-0789.

There are different Facebook groups to help users find contact information. In these groups users may ask recommendations about a product or service to other users in the same group, where users share their experience through comments. Although these groups contains thousands of contact information, search recommendations in these groups is a tedious and time-consuming activity, mainly because the large amount of information contained in this group and the redundant post and comments. To help users find post/comments of interest, Facebook provides a search engine that contains different kind of filters. However, these filter options are limited and sometimes insufficient for those who need to find specific information. This article proposes the use of emotion analysis and machine learning to improve the search for contact information for Facebook publications. This paper also reports an experiment conducted with 15 users, where they asked to search for contact information using the prototype developed and the search engine that provides Facebook. The result shows that most users who used the Facebook search engine find relevant information in the 6th and 7th position of the list of results, while using the prototype developed the user finds the desired information in the 1st or 2nd position of the list.

Keywords : social network; machine learning; naive Bayesian network; classifier; sentimental analysis.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )