This project required building a solution for continuous monitoring of web search systems:
- testing of web search results in terms of quality
- tracking results changes, for example after algorithms change
During project we implemented monitoring of following parameters:
- navigational search quality
- vertical topics search quality
- citation search quality
- Porn sites in response to neutral queries
- Change analysis in results for high-volume queries
Search changes are verified by regular measurement of defined parameters and their comparison with previous measurements.
System was designed to support following search engines: in-house search, Rambler.ru, Yandex.ru, Google.com, Nigma.ru, Bing.com, Yahoo.com .
System was implemented using Python/Django.
Besides the core system other development also included:
- user interfaces.
- list of markers for each type of analyzed queries.