PredictionIO Unveils Open Source Predictive Analysis Suite for E-Commerce

PredictionIO Unveils Open Source Predictive Analysis Suite for E-Commerce

PredictionIO has released its first suite for predictive data analysis, which is available for free under an open source license.

Machine learning startup PredictionIO hopes to use open source to undercut the competition in the data analysis and predictive technology market. The company has unveiled its first suite of product recommendations and other similar tools, all available at no cost and with full source code.

Predictive data analysis technology is already widespread from companies such as RichRelevance, Baynote, ExactTarget and Sailthru. But PredictionIO said the closed-source nature of existing solutions limits their effectiveness for enterprises.

That's why the company, which received $2.5 million in seed funding last July, believes its open source approach to predictive analysis will allow it to make major inroads in a market that is already saturated. Companies of all stripes and sizes "now have access to Amazon-style technology, at no cost," PredictionIO said in a statement.

The company's product release includes five main features for e-commerce sites and related outlets: Personalized product recommendations for shoppers, predictions of the likelihood that a particular visitor will make a purchase; product rankings tied to a customer's preferences; recommendations of similar products (i.e., "You may also like...") and predictions for which products a customer will wish to buy next after making a purchase.

PredictionIO's basic platform is available for free. The company also sells a premium enterprise version with professional support services.

In addition to the offer of a predictive analytics package at no cost, PredictionIO is also pitching the customizability of its code. "With PredictionIO’s open source engine, anyone can build and deploy predictive e-commerce and m-commerce applications in a fraction of the time," the company said. "Developers can further customize the code to fit their unique need."

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.