Boost Genetics

This implements a proof of concept genetic optimization for Apache Solr boost factor optimization problems. A rough outline for that can be found the related blog post.

Running the application

• Install NodeJS NodeJS
• Run 'npm install'
• Run node app.js
• Open the application in your browser-window

Field usage

• Host: The host where Solr can be found (e.g. localhost)
• Port: The port where Solr can be found (e.g. 8080)
• Path: The base path to the select where handler, the query parameter has to be the last one (e.g. /solr/rfcs/select?defType=dismax&rows=50&fl=id&q=)
• Fields: The query fields which should be optimized, actual factors replaced by placeholders (e.g. &qf=title^0%20keywords^1%20abstract^2%20sec1^3%20text^4%20titles^5%20references^6%20author^7)
• Iterations: The amount of generations to run the optimization through - 50 seems to be a good start
• Domain: The allowed value range for the factors - must match to the amount of placeholders in your fields string (e.g. [[1,100],[1,100],[1,100],[1,100],[1,100],[1,100],[1,100],[1,100]])

Licence:

Apache License, Version 2.0 (see LICENSE)