In this project we try to find a similarity measurement for two given locations. We use two datasets from zenodo:

• https://zenodo.org/record/3748719
• https://zenodo.org/record/1001662

The dataset repo is located here: https://github.com/airdocs/datasets.

The utils.py file reads and loads data from dataset repo.

costincarabas@MacBook:~/facultate/doctorat/research/AirDocs$ tree -L 2
..
├── datasets
│   ├── DISTRIBUTED_OPENSOURCE_version2
│   ├── README.md
│   └── UJI_LIB_DB_v2.2
└── similarity
    ├── README.md
    ├── __pycache__
    ├── similarity.py
    └── utils.py

Step1: * Requires whitelist.txt * Outputs p_6-redmi-17-03-2021_12-52-29.json

`python3 preprocessing.py 6-redmi-17-03-2021_12-52-29.json`

Step2: Call compare_locations() function from compare_locations.py file

The similarity_upb.py script: similarity_collection_vs_all(json_file, collections, index = 0, method = 'Average'))

Returns the closest neighbors that have the similarity under **0.1**.