Location Based Social Network Service (LBSNS)

Network Data Analysis CW 2

Multi-layer networks

Package

1. pandas
2. matlibplot
3. xml
4. smopy
5. sklearn

Data source

1. NYC Foursquare Checkins

1. Foursquare Categories Resource

Preprocessing

1. Download check-ins data from Foursquare.
2. 01_data_insight.py
   1. Extract all venueCategory
   2. Count user venueCategory visit frequency.
3. Download official venue category from Foursquare.
4. 02_categroy_extract.py
   1. Extract category from html file.
   2. Map all venueCategory in check-ins to the parent category
   3. If cannot find in official category, then manual input
5. 03_user_loc_cate_mapping
   1. aggrate user check-in behaviou on venue categories
6. 04_food_location
   1. Select out the venues with parent categroy 'Food'
   2. Manually add a self-defined tag to each food venue type
7. 05_visualisation and 06_recommend
   1. test the result of training
   2. visualise results

Recommender System

1. visualisation and recommender
   1. classes and functions for visualise multilayer network
   2. recommender will use one user id and a location to recommend a list food venues
2. test_recommender
   1. select user Id
   2. select one living-community location and one out-of-living-community location
   3. fit into recommender
   4. plot results in map