This stage is part of Fusion’s Machine Learning framework. It is highly experimental and subject to change.
The Machine Learning framework provides tools to evaluate alternative methods of computing and displaying search results. For example, an experiment may consist of a system which has two or more different query pipelines running, and users are randomly served search results using some variant of the system via instrumented pages that capture user response. In such a system, an Update Experiment stage would be used to feed these responses directly back into the running experiment.
The Experiment Update stage checks each document for an experiment ID, a variant ID and a value and updates the experiment accordingly. A default experiment ID can be provided if none is found in the document. By default these documents are silently discarded after processing, but the stage can be configured to forward them down the pipeline to the next stages.
When entering configuration values in the UI, use unescaped characters, such as