Crossing Minds State-of-the-Art Recommendation API

Recommendation and personalization are simple ways to increase sales by making product suggestions to customers based on historical behavior. This can be applied not only to e-commerce but also to music, gaming, events, articles and more.

Crossing Minds is democratizing access to recent gains in artificial intelligence research by building the world’s best content recommendation platform.

Get familiar with our API endpoints and explore their features.

Introduction

The Crossing Minds API follows the REST conventions for resource-oriented URLs. The API accepts and returns either JSON-encoded data or binary-serialized python objects, and uses standard HTTP response codes, verbs, and JWT authentication.

Table of Content

• Low Level HTTP Walkthrough
• API Entities
• DIY API Tutorial - MovieLens
• Authentication
• Items and Users Properties
• Flexible Identifiers
• Business Rules and Scenarios
• Reranking on Item Property
• Filtering on Item Property
• Filtering on User Property
• Error Handling
• API Endpoints
• Under the Hood

Machine Learning Concepts

Machine Learning Concept	Description
Deep Collaborative Filtering	Pinpoint complex patterns to predict taste, in ways that go beyond matrix factorization
Deep Content Extraction	Extract deep information from images or texts to generate meaningful recommendations of new items
Semantic Graph Embedding	Leverage all available information to further increase the accuracy of the recommendations, enabling users to discover hidden gems