All Products
Search
Document Center

Artificial Intelligence Recommendation:AIRec overview

Last Updated:Jun 17, 2026

Artificial Intelligence Recommendation (AIRec) is a personalized recommendation service for enterprises and developers, built on Alibaba's big data and AI technologies. It delivers tailored item, content, and news recommendations to end users without requiring you to build a recommendation engine from scratch. Provide the required data and call AIRec API operations to get recommendations.

AIRec draws on proven experience across e-commerce, content, news, live streaming, and social media platforms.

Industry templates

AIRec provides pre-built templates for three industries, each pre-configured with the data signals and recommendation logic common to that industry. More templates will be added based on customer requirements.

E-commerce template

Recommends products based on attributes such as logistics information and sales data. Optimized for driving direct transactions and meeting click-to-purchase ratio targets.

Common apps: Taobao, Tmall, Xianyu.

Content template

Recommends content on sharing platforms based on engagement attributes such as likes and forwards. Supported content types include short text, articles, images, and combinations of these.

Common apps: Taobao Headlines, marketing communities.

News template

Recommends articles and news based on attributes such as author, publish location, and publish time. Prioritizes timeliness, which is critical in news scenarios.

Common apps: UC Toutiao.

Service types

AIRec provides two recommendation service types that match different stages of a user's browsing session.

You may also like

Use this service when a user's browsing intention is unclear, such as on a homepage or product category page. AIRec learns from both long-term and short-term user behavior, runs training tasks to model user interests, and returns a diverse set of recommendations.

Best for: Homepage, product category page.

You may also like

Related recommendations

Use this service when a user's interest is already focused, such as on a product detail page. AIRec first finds dynamically associated items (1/N products or 1/N articles) by analyzing large-scale interaction data, then finds statically associated items based on attribute and feature correlation between those dynamic results.

Best for: Product detail page.

Related recommendations

How it works

  1. Prepare your data and provide the required inputs.

  2. Select an industry template that matches your app.

  3. Call AIRec API operations with the required data.

  4. AIRec returns personalized recommendations ready to display in your app.

What's next

  • Follow the quick start to set up your first recommendation scenario.

  • Explore the API reference for available operations and parameters.