Frontend

[komodo108]

Implement a frontend to interact with the dataset & see visualisations. Perhaps use d3.js to provide for visualisations & react.js.

Task List

• [x] Implement a search bar allowing users to search the dataset by column type (like one of those search bars with a dropdown menu)
• [x] Talk to the backend to get results
• [x] Display them in a sensible way
• [x] Implement custom views for aggregate data - comparing 2 users, seeing clustering, etc.

[komodo108]

Below is my idea on the search interface, what do you think of this @georgw777?

Search Interface

To search through the dataset, we have implemented a permissive search bar to display the results. This search bar allows the user to search using an input box (which controls the actual search query) & a dropdown option list (which controls the mode to search with).

Search methods are as follows:

• Single user: In user mode, a single integer (id) will display movie & genre statistics about the user with that id
• Multiple users: In user mode, a comma separated list of integers (ids) will display the movie & genre statistics of the collection of users including their favourite genre as a group
  • If only 2 users are specified, we additionally show a visualisation of the differences between their movie tastes
• Single movie: In movie mode, a single integer (id) or string (title) will display the average rating of that movie, as well as the number of users who have watched that movie
• Genres: In genre mode, a comma separated list of strings (genres) will show the list of movies with that genre

Advanced tasks such as clustering or visualisations of the dataset as a whole will be given in an additional page. The movie recommendations will probably be added to the single user search mode.