Real-time item recommendations in Amazon Personalize
If your use case or recipe generates item recommendations, after you create a recommender or create a campaign, you can get real-time personalized or related item recommendations for your users.
If your domain use case or recipe provides real-time personalization, such as the Top picks for you use case or the User-Personalization-v2 recipe, Amazon Personalize updates recommendations based on your user's most recent activity as you record their interactions with your catalog. For more information on recording real-time events and personalization, see Recording real-time events to influence recommendations.
When you get real-time item recommendations, you can do the following:
-
If you configured your campaign to return metadata for recommended items, you can specify the columns to include in your GetRecommendations API operation. Or you can specify the columns when you test the campaign with the Amazon Personalize console. For code samples, see Getting item metadata with real-time recommendations. For information about enabling metadata for a campaign, see Item metadata in recommendations. For information about enabling metadata for a recommender, see Enabling metadata in recommendations for a domain recommender in Amazon Personalize.
-
For some use cases and recipes, you can specify a promotion in your recommendation request. A promotion defines additional business rules that apply to a configurable subset of recommended items. For more information, see Promoting items in real-time recommendations.
-
You can filter results based on custom criteria. For example, you might not want to recommend products that a user has already purchased or recommend only items for a particular age group. For more information, see Filtering recommendations and user segments.
Note
If you used a PERSONALIZED_RANKING custom recipe, see Getting a personalized ranking (custom resources).
Topics
How recommendation scoring works (custom resources)
With the User-Personalization-v2 and User-Personalization recipes, Amazon Personalize generates scores for items based on on a user's interaction data and metadata. These scores represent the relative certainty that Amazon Personalize has in whether the user will interact with the item next. Higher scores represent greater certainty.
Note
Amazon Personalize doesn't show scores for domain recommenders or the Similar-Items, SIMS or Popularity-Count recipes. For information on scores for Personalized-Ranking recommendations, see How personalized ranking scoring works.
Amazon Personalize generates scores for items relative to each other on a scale from 0 to 1 (both inclusive). With User-Personalization-v2, Amazon Personalize generates scores for a subset of your items. With User-Personalization, Amazon Personalize scores all of the items in your catalog.
If you use User-Personalization-v2 and apply a filter to recommendations, depending on how many recommendations the filter removes,
Amazon Personalize might add placeholder items. It does this to meet the numResults for your
recommendation request. These items are popular items, based on amount of interactions data, that satisfy your filter criteria.
They don't have a relevance score for the user.
For both User-Personalization-v2 and User-Personalization, the
total of all scores equals 1. For example, if you're getting movie recommendations for a
user and there are three movies appearing the Items dataset and Interactions dataset, their scores might be
0.6, 0.3, and 0.1. Similarly, if you have 10,000
movies in your inventory, the highest-scoring movies might have very small scores (the
average score would be.001), but, because scoring is relative, the
recommendations are still valid.
In mathematical terms, scores for each user-item pair (u,i) are computed according to
the following formula, where exp is the exponential function,
w̅u and wi/j are user
and item embeddings respectively, and the Greek letter sigma (Σ) represents summation over
all items with scores:
Recommendation reasons with User-Personalization-v2
If you use User-Personalization-v2, items the model wouldn't normally recommend include a reason list.
These reasons explain why the item was included in recommendations. Possible reasons include the following:
-
Promoted item – Indicates the item was included as part of a promotion that you applied in your recommendation request.
-
Exploration – Indicates the item was included with exploration. With exploration, recommendations include items with less interactions data or relevance for the user. For more information about exploration, see Exploration.
-
Popular item – Indicates the item was included as a placeholder popular item. If you use a filter, depending on how many recommendations the filter removes, Amazon Personalize might add placeholder items to meet the
numResultsfor your recommendation request. These items are popular items, based on interactions data, that satisfy your filter criteria. They don't have a relevance score for the user.
