Chart of the Month: Relationship Chart πŸ§‘β€πŸ€β€πŸ§‘

Sheila Bugal on October 10, 2021
Customer Experience

Caplena offers many visualization options, for example, the driver chart or the treemap. This blog post is all about the Relationship Chart – although it’s not too good at sticking to just one relationship at a time 😜. Staying in line with its polyamorous spirit, the relationship chart’s tendency to find relationships wherever it looks is actually a very good thing, as it helps us determine which correlations are important, which aren’t, which are strong – or fragile, so that we can design an effective strategy based on these facts, no matter what our agenda is.

Relationship Chart Structure

Models of data are represented by nodes and links, and the strength of a connection between two nodes depends on their interrelationship. Consider the following diagram ⬇️. ‘Bob’ is a node. Bob is interested in the Mona Lisa painting, which is located in the Louvre, which, in turn, is a museum.

Relationship graph example of how Bob's relationships interconnect. It is an example of how AI can categorize feedback.
Source: www.allthingsdistributed.com

Graphs of relationships can be quite simple, such as the one above. In Caplena’s relationship chart, the concept is the same, but instead of individual people, places, or things, Caplena groups responses to surveys or other types of text comments into bulk categories known as codes.

What’s a code?

Let’s say that instead of one sheet of paper with Bob’s name on it, the node of Bob is a stack of 200 sheets, full of people interested in the Mona Lisa painting. It’s the same principle, but since there is now a stack of people that were categorized as having the same interests as Bob. This is known as a code on the Caplena platform.

Review: "The network coverage and service are great but they're just way too expensive compared to other carriers". Overall negative sentiment.
An example of what codes would be assigned to a text comment.

Applications

Understanding how topics relate to each other can come in handy in a variety of ways. Advanced disease research can employ relationship charts to display links between diseases and gene interactions. By examining these links, you can identify patterns in protein pathways that may contribute to a particular disease. Mind-boggling stuff, isn’t it? A sufficient amount of relationship data will even help you predict the future!

Check out the video below to see how a typical relationship chart would look in Caplena.

Relationship Chart in Caplena

Let’s go through one more example of Caplena usability using an NPS survey on mobile carriers. When you hover over what makes the Unlimited Data code, it shows a correlation with being Cheap/Affordable. This means that when people talk about how cheap and affordable the service is, they often also mention that they have unlimited data.

To stay on theme with the NPS survey of mobile carriers above, imagine yourself as the head of the mobile carrier and how helpful these relationships could be to understand what you need to do to improve and where you should invest your resources. You could invest your time & money in a smarter way if you realize that (for example) unhappy customers are often caused by rude customer service – rather than what you first thought was the issue – connectivity.

Chart Features

There are way more features in Caplena when it comes to this graph, though! You can, for instance:

Change the measurement which is used to compute the correlation between two codes…
Set minimum co-occurrences to plot link…
…or exclude selection of codes!

Want to learn more about creating charts and what chart types are available? Learn more here.

Did you enjoy this article? πŸš€ Feel free to suggest more topics and we will do our best to write about it!

βœ‰οΈJust email our Head of Marketing sheila [at ] caplena.com

PSST…. Would you like a FREE trial?

Caplena conducts sentiment analysis using AI to understand how your customers or employees genuinely feel so you can spend less time on the analysis and more time on the results! πŸ†οΈ
Let's Begin!

You may also like: