Case Studies
Client background: A leading provider of Innovative Business Solutions specializing in cloud based CRM solutions, Multi-channel Marketing, Omni Channel Order management, Business Intelligence, and ERP application
Industry: Technology / ISV
Products/Services:Advanced Analytics
Business Challenge
Categorize products based on user comments, description and reviews on various sites
Find relevant words connecting the product
Unearth complex relations between different verbatim comments and reviews
Map product to product category codes
Business Solution
Text pre-processing and removal of special characters
Word Embedding using a python package called Gensim
Created Word2Vec models
Recurrent Neural Networks (RNN) for mapping product and product category codes, with Tensorflow Backend
Business Benefits
Enhanced the predictive accuracy by leveraging the power of Reinforcement Learning
Acquired actionable metrics from customer reviews and comments
Word Embedding coupled with RNN helped handle unseen data
Environment
Python package – Gensim
Recurrent Neural Networks (RNN)
Get Social