apache/predictionio

★ 12,525ScalaAudience · developerComplexity · 4/5LicenseSetup · hard

Mindmap

mindmap
  root((PredictionIO))
    What it does
      ML prediction platform
      API-based queries
    Templates
      Recommendations
      Similar items
      Classification
    Infrastructure
      Hadoop and HBase
      Elasticsearch
      Apache Spark
    Setup
      Source or Docker
      Apache governance

mindmap root((PredictionIO)) What it does ML prediction platform API-based queries Templates Recommendations Similar items Classification Infrastructure Hadoop and HBase Elasticsearch Apache Spark Setup Source or Docker Apache governance

Click or tap to explore — scroll the page freely

Things people build with this

USE CASE 1

Add a product recommendation engine to an app by collecting user events and querying PredictionIO's API for personalized suggestions.

USE CASE 2

Train a classification model on labeled data using a PredictionIO template without building the ML pipeline from scratch.

USE CASE 3

Deploy a similar-items engine so users browsing a product see related items powered by machine learning.

Tech stack

ScalaHadoopHBaseElasticsearchApache SparkDocker

Getting it running

Difficulty · hard Time to first run · 1day+

Requires Hadoop, HBase, Elasticsearch, and Spark, multiple infrastructure components must be running before training a model.

Apache 2.0, use freely in any project including commercial, keep the copyright and license notice.

In plain English

Apache PredictionIO is an open source machine learning framework built for developers and data scientists who want to add predictive features to applications. Rather than building prediction systems from scratch, teams can use this platform to collect user events, train machine learning models on that data, and then query the results through a standard web API. The goal is to make it practical to deploy machine learning in real products without requiring deep expertise in the underlying algorithms. The framework handles several common prediction tasks through pre-built templates. Examples include recommendation engines (suggesting items a user might like), similar-product engines (finding things related to what a user is viewing), and classification engines (sorting inputs into categories). Each template provides a starting point that developers can customize for their specific use case. Under the hood, PredictionIO relies on well-known open source data infrastructure tools including Hadoop, HBase, Elasticsearch, and Spark. This architecture is designed to handle large amounts of data and scale as usage grows. Installation can be done from source code or via Docker containers. The project is part of the Apache Software Foundation, which means it follows Apache's open governance model. Bug reports and feature requests go through Apache's JIRA issue tracker, and there are mailing lists for both users and contributors who want to follow development or get help.

Copy-paste prompts

Prompt 1

Walk me through setting up Apache PredictionIO with Docker and deploying the recommendation engine template to suggest products to users.

Prompt 2

I want to send user click events to PredictionIO and then query it for top-5 recommendations for a specific user. Write the code for both the event ingestion call and the query request.

Prompt 3

My PredictionIO recommendation engine returns poor results. What factors in the training data or template configuration should I tune to improve accuracy?

Open on GitHub → Explain another repo

← apache on gitmyhub — every repo by this author, as a profile.

Verify against the repo before relying on details.