inqicon

START A CONVERSATION

Share your requirements and we'll get back to you with how we can help.

Thank you for submitting your request.
We will get back to you shortly.

ContextIQ Recommendation Engine

Project Overview

Keeping visitors hooked with personalized content increases the time spent on a site and chances of a conversion. Our client envisioned a solution that would take into account browsing behavior of visitors and serve content that is most relevant to their interests.

Client Profile

Trusted by over 15 million websites, our client is one of the largest social infrastructure and analytics platforms on the web. The company provides easy-to-use consumer engagement tools and services to advertisers, publishers, and brands.

QBurst Solution

Our solution was an easy-to-deploy personalization system called ContextIQ that uses collaborative filtering algorithms to deliver content recommendations. The solution analyzes the web history of users and uses a custom-built URL classifier to predict user interests.

There are well-defined plugins on the client side that record the required behavior. The recorded behavior is transferred to ContextIQ via web service calls and is continuously processed by its algorithms. These algorithms store results onto a Mongo instance. The recommendation engine, which runs on a Hadoop cluster, then returns a set of recommendations based on defined parameters.

Highlights

  • Proprietary algorithm to compute word weight in a bucket
  • Designed to stem top words pagewise and match against buckets
  • Easily deployable and customizable solution
  • Admin interface to choose algorithm and tune parameters
  • Custom predictions based on specific business needs
  • Highly scalable system
  • Mobile friendly

Technologies

  • HBase as columnar data storage
  • Java
  • MongoDB
  • ContextIQ (Proprietary Recommendation Engine)