The Customer Analytics Tool has 3 components

  • Tweets analysis
  • Customer Reviews Analysis
  • Individual Text Analysis

Tweets Analysis

The Tweets analysis can be done for

  • all the Twitter handles
  • Donald Trump handle
  • Hillary Clinton handle

Women's clothing reviews

The Women's clothing reviews analysis can be done for

  • all the Dress types
  • for a particular dress type (`Dresses, Pants, Swim, Jeans, Blouses, Intimates)

The analysis for both the components has the following functionalities

  • Distribution of sentiments
  • Distribution of emotional traits
  • Distribution of behavioral traits
  • Most Important Concepts
  • Most Common Words
  • Most negative reviews

The analysis can be further filtered by

  • Sentiment Category
  • Emotions
  • Behavior

Individual Text Analysis

Individual text analysis of a text is done and the following are shown

  • Overall Sentiment score of the text
  • Emotions in the text
  • Behavior in the text
  • Key Phrases
  • Key Entities
  • Key relations

Expert APIs used in the application

The APIs which have been used are as follows:

  • Sentiment API
  • Emotional Traits API
  • Behavior Traits API
  • Phrases API
  • Entities API
  • Relations API

The apis have been used extensively in the code in the files

  • code / womenreviews-clean.ipynb
  • code / womenreviews-clean.ipynb
  • code / textanalytics.py

How we built it

This was built with Expert ai API, Python, and Streamlit

Challenges we ran into

At first, we were limited to the number of API calls in the first few days. But the Expert AI team removed the limitations and I was very pleased to see the quick reaction time.

Accomplishments that we're proud of

We built a functional application and a platform for customer sentiment analysis comprising of tweets analysis, customer brand reviews analysis, and specific customer input analysis

What we learned

We learnt about the Expert AI APIs and how easily we could integrate with Streamlit . This was amazing and I could deploy the application in Azure , AWS using Azure Containers , Azure Kubernetes Service , Heroku and AWS Fargate. This was amazing learning experience with Expert AI

What's next for Customer Sentiment Analysis Tool

The Customer Sentiment Analysis Tool is like a customer sentiment platform. This can be extended with further functionalities of real time analysis. Moreover, we were succesful in deploying on premises and in various cloud providers, so it was very amazing

Built With

Share this project:

Updates