In addition to the basic search experience enabled through query pipelines, Fusion provides ways to develop an enhanced search experience for your end users and provide useful data for your analytics team. The primary mechanisms for doing this are signals and aggregations.
By collecting signals and aggregating them, you compile a body of data that allows you to develop a sophisticated search experience, with rich search results for your end users, based on past user behavior.
Signals and aggregated signals are stored each in their own collection. These collections are associated with a primary collection, so that a collection named "products" will have two related collections: "products_signals" and "products_signals_aggr". By default, when using the UI to create a collection, a "signals" and "aggregated signals" collection are also created.
See also these subtopics:
Signals are events that are collected for analysis or to enhance the search experience for end users. Common types of signal events include clicks, purchases, downloads, ratings, and so on.
You can use App Insights to get visualizations and reports with which to analyze your signals data. App Insights mainly uses raw signals, but also uses some aggregated signals.
Aggregations are processed signals. An aggregator reads the raw signals and returns interesting summaries, ranging from simple sums to sophisticated statistical functions.
Crucially, it must be possible to relate the documents in an aggregated signals collection to documents in the primary collection, in order to use the aggregated signals for recommendations and/or boosting of searches over the primary collection.
The cold start problem
The "cold start" problem means it is hard to personalize the search experience when insufficient signals have been aggregated. For example, it is hard to offer recommendations to users who have never visited before, or for queries that have never been issued before, or for items that have been recently introduced into the system.
Fusion provides solutions for this problem using its query pipelines. A query pipeline that includes stages for blocking, boosting, or recommending based on signals can also include stages that provide fallbacks. In the case where there is not enough data to provide specialized blocking, boosting, or recommendations, the pipeline can return a simpler set of search results using Solr’s normal relevancy calculation.
A common solution to the cold start problem is to sort or boost on a certain field to provide pseudo-recommendations when more specific recommendations are not available. For example, you can sort on the
sales_rank field to recommend the most popular products, or boost on the
date_added field to recommend the newest items.