Behaviour Analysis System

Our Experience developing a behaviour analysis System integrated with AI and ML for user interest’s identification.

Customer Profile:

Our client is a premier global provider of market intelligence, advisory services, and events for diverse verticals based out of Japan. With the mission to help people turn their curiosity into action, they enable curious individuals and companies – including 98% of the Fortune 500 – to have conversations at scale with the people who matter most.

Problem Statement:

Although traditional mode of marketing like television and newspapers reach masses and potential customers, it is less effective due to high cost and low conversion rates. From a consumer perspective, they receive lot of unnecessary promotions and irrelevant advertisements. To tackle this, recommender systems is employed to identify user interests.

A survey solution is a feasible way for user profiling without infringing user privacy. Hence, Innovature developed a user interests identification solution with self-service questionnaire for user profiling and to serve them accordingly.

Project Background:

Being a keen observer of the current technology trends and changing customer requirements the client wanted a self-service questionnaire software that analyses user behaviour and interests through the questions answered by the user.

There are two main techniques to achieve this: (i) Content based and, (ii) Collaborative filtering. Content based technique matches user interests directly with the kind of solution they need. Collaborative filtering suggests additional services which may not be of direct user interest but derived through analysing a bunch of people with similar interests. To illustrate this, a company finds out that people who buy fish normally buys rice as well. Now the company will recommend rice to a fish buyer even if that user has never bought rice in the past. It is ideal to combine both Content based filtering and Collaborative filtering to achieve the merits of both systems.

Challenges:

  • Implement a data crunching and analytical subsystem to the new platform.
  • iPad application with web server interaction using WebView actions

Technology:

Development language: Ruby on Rails

Database Server: MySQL

File Storage: AWS S3 Hibernate

Content Delivery: AWS CloudFront

Solution:

Innovature built a web questionnaire solution that generates questions intuitively based on user responses and thus analyses the interests and behaviour patterns of the user. Additionally, the solution also protects the privacy of the user as no data is collected in the background. The user himself gives answers. The results are then analyzed individually and with other people’s list who have taken the questionnaire to arrive at meaningful insights.

Achievements:

  • Convenience
  • More control

Return on Investment:

The user interest identification app accounted to 130% improvement in customer’s survey performance.