Relative Insight's comparative approach is what makes our technology unique from other text analysis solutions. In this article, you will learn:
What is comparative text analytics?
Comparative text analytics uses principles of comparative linguistics to reveal the statistically significant differences and similarities between two or more text data sets. In other words, it helps you understand what makes a particular data set (and the audience group, brand or product it represents) unique and similar to others.
How is comparative text analytics different from other types of text analysis?
The text analytics capabilities embedded in social listening, survey analysis and customer insight tools often rely on pure frequency-based analysis. This type of analysis reveals the dominant words and phrases in a particular data set and are often presented in the form of a word cloud.
There are two primary challenges with this approach:
It often surfaces obvious discoveries (of course people who talk about the supermarket use the word shopping).
It fails to incorporate the necessary context into the analysis.
While it may be informative to know that 50 customer reviews include mentions of the phrase 'great quality', this information is not insightful without context. To truly understand if this is something to be celebrated, we need answers to some other questions, for example:
Is 50 mentions more or less than the competition?
How does it compare to reviews from last year?
Are the comments specifically related to one product?
Relative Insight's comparative approach provides answers to these kinds of questions by incorporating reference points into the analysis.
What are the benefits of comparative text analysis?
There are three primary benefits to the comparative approach.
It incorporates context into your analysis which helps to generate more robust insights that hold up under scrutiny.
It surfaces only the statistically significant differences and similarities between data sets, helping users overcome the hypothesis-driven bias that is inherent in human-led (manual) analysis.
It enables the scalable analysis of text data by focusing your attention on the things that matter within your data and nothing else.
Let us explain!