E-Commerce Intelligent Recommendations

This solution helps enterprises build precise user models and provides consumer-specific recommendations based on in-depth analysis of consumer behavior and transaction history.
Solution Advantages
  • Real-Time Recommendations

    Uses a leading online recommendation algorithm to return consumer-specific recommendations at ultra-low latency, specifically, in milliseconds (< 100 ms).

  • Feature Profiles from Tens of Millions of Dimensions

    Uses high-performance distributed recommendation algorithms to analyze the correlations between user behavior based on multiple vectors, draws feature profiles from tens of millions of dimensions, and enables minute-level model updates.

  • Automatic Data Model Updates

    Automatically discovers data features and dynamically adjusts data models based on information changes in real time, for more accurate recommendations.

  • Effectiveness Evaluation

    Provides historical recommendation records and individual user profiles, allowing carriers to analyze the product experiences of specific types of users to find opportunities for improvement.

Business Challenges
  • Low Conversion Rates

    Traditional marketing strategies focus on categories of consumers rather than being consumer-specific. E-commerce platforms recommend the same products to all consumers, leading to low conversion rates.

  • Poor User Experience

    Consumers accessing online stores are presented with the same set of products, regardless of personal preferences.

  • Excess Inventory

    Ineffective marketing strategies lead to low conversion rates. As a result, there is a continuous increase in unsold inventory.

  • High Cost

    E-commerce platforms need to invest heavily in recommendation engines that help implement precision marketing.

Solution Architecture

Solution Description

This solution helps enterprises build precise user models and provides consumer-specific recommendations based on in-depth analysis of consumer behavior and transaction history.

Architecture Description

The Computing Engine uses machine learning and deep learning algorithms to meet complex recommendation needs.
The Recommendation Engine provides a complete framework (from data input and computation, to data output), and allows customization of algorithms to meet specific business needs.
The User Profile component leverages consumer favorites, behavior, transaction history, website and mobile app usage, location, logs, and other data to generate 360-degree user models.
API provides external users with RESTful APIs to invoke services, easily interconnecting recommendation systems with existing business systems.

Recommended Services
  • MapReduce Service
    MapReduce Service (MRS) provides enterprise-level big data clusters on the cloud. Tenants can fully control clusters and easily run big data components such as Hadoop, Spark, HBase, Kafka, and Storm.
  • Machine Learning Service
    Machine Learning Service (MLS) helps you quickly find data patterns to construct prediction models through machine learning technologies and deploy these models as prediction and analysis solutions.
  • Deep Learning Service
    Deep Learning Service (DLS) is powered on the high-performance computing capabilities of HUAWEI CLOUD. With built-in optimized network models, DLS allows you to implement model training, evaluation, and inference with the flexibility of on-demand scheduling.

Create an Account and Experience HUAWEI CLOUD for Free

Register Now