Practical Machine Learning Innovations in Recommendation[搬运]

410234198

3129

收藏 2014-08-06

by:Ted Dunning, Ellen Friedman
http://www.oreilly.com/data/free/machinelearning.csp?intcmp=il-strata-free-product-lgen_machinelearning

Print Length: 53 pages
It`s a short e-book.

Building a simple but powerful recommendation system is much easier than you think. This report explains innovations that make machine learning practical for business production settings—and demonstrates how even a small-scale development team can design an effective large-scale recommender. The style of the report makes this subject approachable for all levels of expertise.

Authors Ted Dunning and Ellen Friedman walk you through a design that relies on "careful simplification." You’ll learn how to collect the right data, analyze it with an algorithm from the Apache Mahout library, and then easily deploy the recommender using search technology with Apache Solr. This powerful and effective combination is efficient: it does learning offline and delivers rapid response recommendations in real time.

Understand the tradeoffs between simple and complex recommenders
Collect user data that tracks user actions—rather than their ratings
Predict what a user wants based on behavior by others, using Mahout for co-occurrence analysis
Use Solr to offer recommendations in real time, complete with item metadata
Watch the recommender in action with a music service example
Improve your recommender with dithering, multimodal recommendation, and other techniques

contents：

1. Practical Machine Learning
  What’s a Person To Do?
  Making Recommendation Approachable
2. Careful Simplification
  Behavior, Co-occurrence, and Text Retrieval
  Design of a Simple Recommender
3. What I Do, Not What I Say
  Collecting Input Data
4. Co-occurrence and Recommendation
  How Apache Mahout Builds a Model 16
  Relevance Score
5. Deploy the Recommender
  What Is Apache Solr/Lucene?
  Why Use Apache Solr/Lucene to Deploy?
  What’s the Connection Between Solr and Co-occurrence Indicators?
  How the Recommender Works
  Two-Part Design
6. Example: Music Recommender
  Business Goal of the Music Machine
  Data Sources
  Recommendations at Scale
  A Peek Inside the Engine
  Using Search to Make the Recommendations
7. Making It Better
  Dithering
  Anti-flood
  When More Is More: Multimodal and Cross Recommendation
8. Lessons Learned
A. Additional Resources