This section focuses of the relevance tuning phase of your search project.
Coveo offers powerful tools to help you continuously improve relevance in your search solution.
Most interactions with your search experience record usage analytics (UA) data in your Coveo organization. By regularly inspecting dashboards and reports based on this data, you can identify relevance issues in your search solution. You can then configure query pipelines to conditionally apply ranking expressions, thesaurus rules, filters, and other manual relevance tuning features to deal with these issues.
Once you have gathered enough UA data and your search solution is yielding satisfyingly relevant results, you can generate Coveo Machine Learning (Coveo ML) models to start providing your users with increasingly relevant and personalized interactions.
This is one of the most important sections of the Coveo project guide. Previous sections should be considered prerequisites. Tuning relevance on top of a messy index or badly designed search experience will likely have little impact. For more information on these topics, see:
Articles in this section