TÉCNICAS DE MACHINE LEARNING APLICADA A MINERAÇÃO DE DADOS E ANÁLISE DE SENTIMENTOS PARA PREDIÇÃO DE HOMOFOBIA NO TWITTER

نویسندگان

چکیده

Este trabalho estuda a identificação de tweets homofóbicos, utilizando uma abordagem processamento linguagem natural e aprendizado máquina. A metodologia mineração dados aplicada neste foi CRISP-DM. O objetivo é construir um modelo preditivo que possa detectar, com razoável precisão, se Tweet contém conteúdo ofensivo indivíduos da comunidade LGBTQIA+ ou não. banco utilizado para treinar os modelos preditivos construído partir diversos coletados. Foram coletados mais 3000 desenvolvimento do nosso trabalho, obtendo resultados 86% precisão.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Uso de instâncias de dados e carga de trabalho para mineração de restrições de integridade

Functional dependencies (FDs) are integrity constraints widely studied in the context of data profiling. In this work, we explore the automatic discovery of FDs and describe a method for selecting relevant ones regarding workload semantics. The experimental evaluation shows that the selected dependencies exhibit expressive properties compared to the search space, which demonstrates the effectiv...

متن کامل

Análise de Padrões de Propagação no Twitter

As a consequence of the popularization of the social networks, analyzing the information propagation in such networks has become a relevant task in several scenarios. Propagation patterns support the understanding of phenomena such as opinion formation and the emergence leaders in social networks. In this paper we present a methodology for propagation analysis on Twitter, the most popular micro...

متن کامل

Uma Avaliação de Eficiência e Eficácia da Combinação de Técnicas para Deduplicação de Dados

Data Deduplication is the task of identifying and eliminating duplicate records in a single database. It is a complex process that involves several steps, including: defining blocking key, similarity function and indexing method. There are several approaches for each of these steps. In this context, the objective of this work is to find the best combination for such algorithms aiming to improve...

متن کامل

Análise Experimental de Bases de Dados Relacionais e NoSQL no Processamento de Consultas sobre Data Warehouse

Data warehouse (DW) is a large, oriented-subject, non-volatile, and historical database, and an important component of Business Intelligence. On DW are executed OLAP (Online Analytical Processing) queries that often culminate in a high response time. Fragmentation of data, materialized views and indices aim to improve performance in processing these queries. Additionally, NoSQL (Not only SQL) d...

متن کامل

Análise de interdependência dos habilitadores tecnológicos, empresariais e humanos no desenvolvimento de base de dados pessoal

The main subject of this study is the development of an environment integrating personal data and digital services (e-services), in which we are denominating of Personal Data Base Solution. The solution harmonizes and aligns the personal, business and social interests, through the several specific benefits for each one of these entities, obtained by the common and integrated solution. Besides d...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Foco

سال: 2023

ISSN: ['1981-223X']

DOI: https://doi.org/10.54751/revistafoco.v16n1-121