This page describes the recommendations types that are available when you create a media recommendations app, including the business objectives, context event types, and other customizations that are available with each recommendations type.
Recommendations types
The following recommendations types are available with media recommendations apps:
Others You May Like
The Others You May Like recommendations type recommends the next document that a user is most likely to engage with. The recommendation is based on the user's engagement history and the current context document.
Default optimization objective: click-through rate
Default serving config: N/A
Available customizations:
- Business objective
Supported pages for app deployment:
- Detail page. For more information, see the
view-item
andmedia-play
events.
User event requirements:
view-item
ormedia-play
for click-through rate objectivemedia-complete
and eithermedia-play
orview-item
for conversion rate objective and watch duration per session objective
Recommended for You
The Recommended for You recommendations type predicts the next document that a user is most likely to engage with based on the engagement history of that user. This recommendations type is typically used on the home page.
Recommended for You can also be useful on category pages. A category page is
similar to a home page, except that you display only items from that category.
You can achieve this using a standard Recommended for You recommendations type
with filter tags. For example, you can add customized filter tags (corresponding
to each category page) to the items in your data store. When you send the
recommendation request, set the user event object as view-category-page
and
specify a specific category page's tag in the filter field. Only recommendation
results matching the requested filter tag are returned. Diversity should be
disabled in this use case, because diversity can conflict with category-based
filter tags.
If you use Recommend for You on the home page, select Home Page Context as the
context event type for the app. Home Page Context can deliver better performance
because it uses additional usage signals that are not available with the Generic
option. Only use the Generic option if you don't have view-home-page
events.
Default optimization objective: click-through rate
Default context event type: Home page context
Default serving config: N/A
Available customizations:
Business objective
Context event type
Supported pages for app deployment:
- All. On category pages, you must provide filter tags.
User event requirements:
view-item
ormedia-play
for click-through rate objective. For home page context,view-home-page
is also required.(
media-play
orview-item
) andmedia-complete
for conversion rate objective and watch duration per session objective. For home page context,view-home-page
is also required.
More Like This
The More Like This recommendations type recommends media that is similar to a context item and is likely to be engaged with next by a viewer of the context item. More Like This is based on the context item and the aggregate viewing history of all users who viewed the context item. A More Like This app is typically used on detail pages or on the home page with a fixed context item.
The More Like This recommendations type uses a variety of factors to determine
how similar two media documents are, including the categories field of the
media documents. For best results, media documents that are similar should have
overlapping categories—for example, ["Action", "Comedy"]
categories are
somewhat similar to ["Action", "Thriller"]
, but not similar to ["Drama"]
.
Default optimization objective: click-through rate
Default serving config: N/A
Available customizations:
- Business objective
Supported pages for app deployment:
Detail page
Home page (requires a context item)
User event requirements:
view-item
ormedia-play
for click-through rate objective(
media-play
orview-item
) andmedia-complete
for conversion rate objective and watch duration per session objective
Most Popular
The Most Popular recommendations type recommends media that has been most popular among all users in recent days. The recommendation is based on the watching or viewing history of all users. You can customize the time window to check the popularity of the documents.
Default optimization objective: click-through rate
Default serving config: N/A
Available customizations:
Business objective
Time window. Specify the time window in days to check the popularity of documents in the last X days.
Supported pages for app deployment:
- Home page
User event requirements:
view-item
ormedia-play
for click-through rate objective(
media-play
orview-item
) andmedia-complete
for conversion rate objective
Limitations:
- Most Popular doesn't support customizing serving configs or creating multiple serving configs
Optimization for business objectives
The machine learning models that underlie media recommendation apps are created to optimize for a particular business objective, which determines how the model is built.
After you have trained an app (which is training the underlying model), you cannot change the optimization objective. You must train a new app to use a different optimization objective.
Recommendations for Media supports the following optimization objectives.
Click-through rate (CTR)
Optimizing for CTR emphasizes engagement. You should optimize for CTR when you want to maximize the likelihood that the user interacts with the recommendation.
CTR is the default optimization objective for the Others You May Like and Recommended for You recommendations types.
Conversion rate (CVR)
Optimizing for conversion rate maximizes the likelihood that the user consumes the content up to the conversion threshold defined in the app.
The conversion threshold can be specified in seconds or percentage. For example, if the conversion threshold is set to 25% and the user watches at least 25% of the program, then the conversion objective is met.
Watch duration per session
Optimizing for watch duration per session maximizes the duration of media
consumption. This objective uses information from clicks, conversions, and watch
duration derived from media-complete
user events in order to recommend
items that have a higher probability of being watched longer than others.
When logging user events, make sure that the mediaProgressDuration
is
non-negative and is correctly logged in media-complete
events in the last 90
days.
The watch duration per session objective is available for Others You May Like, Recommended for You, and More Like This recommendation types.