Customer Reviews Visualisation
Interactive Data Visualisation
Overview
Data visualisation for Bitext of their linguistic analytics service as an interactive tool.
Customer: Bitext - A Deep Linguistics Analysis company from Madrid, Spain
Involvement: Product Design (UX & UI) & Interaction Design.
Problem
Bitext's service analyses text and automatically extracts several attributes from the text, such as entities and concepts, as well as additional information that allows them to categorise sentences and paragraphs by their primary topic.
Bitext also provides sentiment analysis by identifying the topic of the opinion and detecting which attributes or features of the text are seen as positive or negative. As an alternative to a sentence-by-sentence, phrase-by-phrase visualisation, Bitext was keen to illustrate how each piece of information present in the text is related to each other and how it is summarised in a visually appealing and easy to understand way.
Solution
To be able to represent both relationships within the text and the number of times a particular attribute was present we chose a node graph. Given the large amount of data we faced, the issue was making a graph that would be functional and easy to understand, while still representing all the information.
I explored adding a search for words within the text and the function to explore the levels of connection from one node. In addition I added a bar chart for the totals for each topic. I added ability to collapse down the sidebar to allow a user to explore the graph with full screen. Perfect for screenshots for presentations.
Concepts Visualisation
Sentiment, Entities & Categories Visualisation
Result
In total we created four visualisations for each of the different groups and showed it with English and Spanish written reviews. The client was delighted with final interactive visualisation.
