create Astro Ratings Matrix
cARM is a software tool that creates a dataset suitable for evaluating recommender systems for Open Cluster of Stars using Scientific Literature.
- python 3.5+
- sqlite
- numpy
- pandas
- configargparse
-
The first step is to create an account at ADS - follow the instructions in https://github.com/adsabs/adsabs-dev-api to create your ADS API token
-
Second, you need to create the sqlite database - the commands are in createDB.sql
-
Third, make sure to complete the info at config.ini file, including the correct paths to the database and clusters file
- Now you need to fill the database with the information about the clusters:
python fillClustersTable.py
- Next step is performed to find all the articles for each cluster in our database
python fillTables.py
cd authorClusterMatrix/
python createAuthorClusterMatrix.py
cd authorClusterMatrix/
python createAuthorClusterMatrix_seq.py
The script /clustersNameAnalysis/clustersNameAnalysis.py returns groups of names of clusters whose simbadID was not found
For download of an example of dataset created with this software: https://drive.google.com/drive/folders/17a5rSq1iIy4vsgAuj0TGHTY3UbaVpvqq?usp=sharing
Here is a simple script to create the subsets of the full dataset, by removing users with a minimal number of rated items: https://bitbucket.org/snippets/mcabarros/aenGGo/remove_users_by_min_rating