Text Mining

Sentiment analysis & topic modelling

The Problem

Most companies do very little with their text data, as it's difficult to analyse and measure quantitatively.

How can you extract sentiment, topic and meaning from your text data and mine valuable information locked up inside the sentences?

The Solution

We use the leading techniques in Natural Language Processing to build models that gauge the sentiment of your customer feedback, and categorise by topic.

Our NLP system tracks how customers are reviewing the various aspects of your business over time and flags if a particular area is starting to receive more negative reviews than usual, so that you can fix the problem immediately.

We integrate this solution with your social media platforms such as Twitter and Instragram, so that incoming feedback is instantly analysed and converted into meaningful, actionable information.

Inputs

Text data of any kind (surveys, reviews, documents etc)

Deliverables

1. A model that predicts the sentiment of a passage of text, returning a number between 0 and 1

2. A model that extracts key phrases from a passage of text, that capture the topic or subject of the response

3. Integration of the models into your social media, chatbots and survey feedback tools, with output fed to a live dashboard that tracks the sentiment of each topic over time

Outcome

You'll be able to truly understand the key factors that are affecting your customers from free text responses and social media feedback, without the need to run tedious and expensive tick-box surveys.

More information