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