Performance analysis of the Oráculo framework for data collection from Twitter / Análise do desempenho do Framework Oráculo para coletas no Twitter

• Autor(a) principal: de Oliveira, Hércules Batista
• Data de Publicação: 2020
• Outros Autores: Guelpeli, Marcus Vinicius Carvalho
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Revista Veras
• DOI: 10.34117/bjdv6n12-549
• Texto Completo: https://ojs.brazilianjournals.com.br/ojs/index.php/BRJD/article/view/22011
• topic: performance analysis; text collector; framework; Twitter.

Online social networks are important social spaces for human interaction, with far-reaching applications in communication, entertainment, advertising, social campaigning and community empowerment. Shared data have become a research source for several studies seeking to analyze user interactions in these networks. Because of the large volume of data produced, text mining techniques are required for analyzing the collected data efficiently. One of the challenges of the text mining process is the lack of direct access to data from online social networks, which requires the use of specialized tools for collecting data. The present study conducts a performance analysis of Oráculo Application Development Framework as a tool for collecting and mining texts shared on the social network Twitter. In this framework, different algorithms and techniques were applied to circumvent the limitations imposed by the Twitter API. Performance tests were conducted comparing the Oráculo and DMI-TCAT algorithms. The results show that Oráculo presents superior performance in the number of tweets collected compared to DMI-TCAT considering the algorithms and scenarios analyzed.