Today, personalization is the basis of all kinds of information websites, e-shops, music and film databases, but also search and other areas. Metarank software, which is completely open source, has recently appeared on GitHub, so developers from all over the world can use it for free.
Metarank uses machine learning to analyze data, but does not require you to master machine learning. Developers can easily connect data via JSON format and then set the necessary parameters via CLI and YAML configuration. The architecture is based on LambdaMART and xgboost technologies, making it an ideal solution for those who do not want to build a personalization system from scratch.
There is also a sample demo (explanation of the implementation), but it is necessary to point out that the project is in the alpha phase, so there are still a lot of bugs to fix. Metarank is available under the Apache 2.0 license.
Metarank’s main developers aim to make the system usable wherever it is necessary to sort and adjust the order of content, both for individual users and other entities. It is thus suitable not only for various software and the web, but also for evaluating data for scientific purposes.