Recommendation engine - Unsupervised Learning to
Discover Hidden Patterns in your Content

Tech Triveni speaker Shubham Goyal

Sriram Sitaraman

Practice Head - Analytics & Data Science

Srijan Technologies

About Sriram Sitaraman

Sriram Sitaraman, with 22 years of international experience in Designing and delivering Innovative and Transformative solutions to diverse sectors like Hi-Tech, CPG & Retail, Manufacturing, Banking, Finance & Capital Markets, Hospitality & Travel, Health Care, Energy, etc. Being an expert in Analytics, Data Science, Business Intelligence, Big Data & Supply Chain, Sitaraman is proficient in Business Intelligence and Information Management technologies. He has expertise in Machine language, Statistical Modeling techniques & Digital Transformation. He is proficient in enabling, mentoring and coaching cross-functional Agile teams. He has also established and headed PMO established Centres of Excellence (COE) and Shared Services units.

With valuable experience in people management and leading high-performance teams, sharing his wisdom through various platforms digitally as well as in a person is just an extension of Sitaram’s role as a mentor within Srijan. Sriram is Decision Maker, Problem Solver, Risk-taker & team leader, all in one. He is Practice Head for Analytics and Data Science at Srijan Technologies.


Recommendation systems are a big part of Online applications because they are related to understanding people. Few well-known implementations that we come across daily are while buying things on Amazon, watching movies on Netflix, browsing through photos on Instagram / Pinterest. Recommendation engines have pervaded every single aspect of our digital life, and they will continue to shape our lives in the future. Content recommendations on web sites are a must nowadays to engage Users / Visitors / Customers spend more time on your site and consume more offerings.

This session plans to demystify Content Recommendations putting it in the context of Natural Language Processing (NLP) and Machine Learning, covering their implementations spanning use and complexity levels - with Similarity implementations and Collaborative Filtering to Matrix Factorization methods. We will deep dive into the different ways of coming up with the Recommendations, what suits a specific need and the Pros & Cons. The session will explain how to implement the Engine using an Unsupervised Machine Learning approach, integration with Drupal, and touch upon the Architectures. Few use cases across domain and complexities will be highlighted.

Share the talk