Our Group
We are a group of four passionate person who want to discovery some
interesting factors about movie tracks.
Yan Liu
Ms in Analytics
Yao Huang
Ms in Analytics
Yuhui Tang
Ms in Analytics
Xuechun Wang
Ms in Analytics
Background
Soundtrack is one of the key parts in the movie, which effectively conveys the theme and emotion of the whole storyline. Good soundtracks could add popularity to a movie while celebrated movies might correspondingly bring attention to its soundtracks.In addition, the soundtrack itself is an individual piece of art that could be described by multiple music characteristics such as tempo and loudness. While in previous studies, researchers have investigated how the soundtracks would affect audience’s emotion (Hoeckner et al.) and the potential relationship between the search volume of movie soundtracks and the movie revenue (Xu & Goonawardene),there are limited researches which study the quantitative features of movie soundtracks.
Overview
In the first part of the project, we web scraped data of popular movies from IMDb. Then we acquired the the corresponding soundtracks from Spotify and collected the detailed music attributes descriptions. We cleaned the data by checking the scope and null values and merged these two datasets into a complete one for analysis.
In the next step in the project, we will conduct exploratory analysis and predictive analysis for the data we cleaned previously. In the part of exploratory analysis, our report consists basic statistical analysis and the insights we derived from data cleaning, histograms on selected attributes and correlations of all variables, cluster and frequent itemset mining analysis on selected attributes. In the section of predictive analysis, we will propose three hypotheses and use the t test and linear regression to verify our hypothesis. In addition to parametric statistical tests, we also use predictive models to evaluate the selected variables on our dataset.
After these detailed analysis using different methods, we will look at our data from different perspectives and have a better understanding on the embedded relationship among variables. The results from analysis will prompt us to investigate the relationship between movies and soundtracks more thoroughly.