Entretenimentos estudos da base de dados audiovisuais

Abstract

With the huge increase in providers of films and television series, such as Netflix, Amazon Prime, Google Play, HBO Go, Net Movies, Itunes Store and Microsoft Store, apart from the traditional movie studios, it is of great interest to discover the most popular tastes. popular with people for this type of entertainment and thereby help them choose based on their characteristics such as type, gender, duration, year, etc. Therefore, it is important to know and analyze a set of data that represents the most diverse types of audiovisual, television and / or cinematographic entertainment, and their variables. Thus, the objective of this work is to put into practice the learning acquired during the postgraduate course in statistics at UFMG, in the exploration, organization and selection of data, as well as to collect a sample for the development of statistical studies learned during the course and thereby making critical analyzes of quantitative and qualitative information, properly interpreting this type of data. The steps were from the extraction and transformation of the data, unifying them in a single file, using the sampling methodology to determine the sufficiently large size of a sample and describing the statistical methods of linear or non-linear regression that well represents the variable of interest (average rating). However, the initial purpose did not offer good adjustments to its models with all the assumptions violated. Thus, we started to study qualitative variables, changing the focus of the study to a parametric and non-parametric analysis of variance (ANOVA) in order to test whether the type of film has any relation with the genre. In addition, a Kruskall Wallis analysis was used to compare the levels of the gender and type variables in their particular form.