Tuning Solr search engine parameters by treating them as an optimization problem and employing a genetic algorithm utilizing normalized discounted cumulative gain [NDCG] as a fitness measure.
-
Clone this repo
-
Run the start_solr script to start Solr
-
Run the following to import the dataset
$ cd dataset $ curl 'http://localhost:8983/solr/restaurantsCollection/update/json?commit=true' --data-binary @restaurants.json -H 'Content-type:application/json'
- Distributed Evolutionary Algorithms in Python - DEAP documentation
- Genetic Algorithm in Adaptive Web Search - wordpress.com blog
- Measuring Search Relevance - eBay Tech Blog
- Discounted Cumulative Gain - wikipedia
- Normalized Discounted Cumulative Gain - kaggle
- Python Solr Module - solrpy