I’m looking for some inspiration from you guys. Based on below description, I welcome all suggestions as to how the most interesting research question, problem formulation, and/or hypotheses might look in your eyes:
I’m doing a Big Social Data Analytics project (university, graduate level) where I’m going to fetch text (posts, comments, comment replies) and engagement (likes, shares, reactions etc.) associated to each post from a case company’s Facebook Page over a 5-year period. I do have unique actor/user IDs as well.
The project is conducted with a large brewery that is currently using its Facebook Page for getting its marketing messages across, but it has never done e.g. sentiment analysis on its posts such as what types of posts receive the most negative sentiment (perhaps do less of this), and what types of posts receive the most positive sentiment (perhaps do more of this). The tool I use for sentiment analysis can analyze both emotions and whether the post/comment is generally positive, negative or neutral. Besides this, I have a tool that can analyze what are the top keywords being used on the brewery’s Page and perform topic modelling out of it.
I also have access to the brewery’s sales data for the same period as well as weather data, if I would think in examining potential correlations between Facebook performance, sales performance and weather (low/high temperature, low/high level of mm rains etc.), and potentially use it for predictive analytics as well.
I’m planning to use nothing more than Alteryx, Excel and Tableau for data preprocessing, and visualization. Looking forward to be inspired with interesting ideas!