New: Text Mining API for customer feedback
Text mining on customer feedback
Want to identify points of improvement and tackle them? Underlined offers a fully automatic text mining solution to analyze feedback flows. With the Text Mining API, the open answer from your customer is enriched per record with the most important topics and sentiment. This gives you insight into what customers say about you and how you can improve your services. The API allows you to visualize the feedback data in a straightforward dashboard. This way, it is immediately clear where your greatest opportunities lie.
Connect to your CX program
Do you want to connect other sources, such as calls, chats, or e-mails, to your customer feedback? Speed up your CX program with the Text Mining API. The API can be used structurally or as a turn-key solution to analyze all your CX management information and insights, via intuitive dashboards to customer experience professionals, data scientists, management and other stakeholders. Data is sent over to the API in a required format so it can be processed. The results are sent back.
How it works
The Underlined Text Mining API provides advanced processing of natural language for plain text. The API connects with a framework of machine learning and AI algorithms, which can be used directly as part of your customer experience management program.
The service offers a wide range of insights:
- Sentiment scoring: Raw text is analyzed for indications of positive and negative feelings.
- Topic detection: Key terms are automatically extracted from the text so that the key points are quickly identified.
- Available taxonomies: Automotive, Banking, Energy, Healthcare, Health Insurance, Insurance, Municipalities, Pension, Publisher, Retail, Telecom. There are always new branches in development.
- The employees were friendly. The waiting time was very long.
- Topic: Attitude & behavior employee.
- Subtopic: Kindness.
- Explanation: Kindness is assigned because:
- The first sentence is given higher priority.
- Friendliness from driver modeling and existing theories appear to have a major influence on customer satisfaction.
With long texts, topics mentioned in the first sentences are given higher priority than topics mentioned later in the text. In texts with multiple topics per sentence, the algorithm prioritizes based on the frequency and the effect the aspects have on customer satisfaction and loyalty. This effect has been established based on driver models developed in-house and existing theories such as Noriaki Kano's three-factor theory.
The Text Mining API is developed together with leading universities. Hence, the API is scientifically based and the reliability is high. Furthermore, the models the API uses are continually trained.
SNS: a satisfied customer
SNS has a couple of strategic themes. Customer retention is one of them. Within this theme, achieving a positive NPS is an important objective. To accomplish this, SNS needed to get more insight into the NPS feedback. By using text mining, SNS was able to find clear indicators for starting the prioritizing process for improving the customer experience, which in turn, helped to increase the NPS.
Underlined started in 2002, focussing on making sense of unstructured data in survey texts and customer reviews, also on social media. Underlined is now the market leader in Dutch language text mining, a service they co-created with the University of Amsterdam. With their customer and data focus, they can uncover CX data insights faster, deeper and better than many of their corporate counterparts. Many clients use their API as an insights-accelerator for their customer experience programs.
Test for free
Are you curious about the Text Mining API, but would you like to try it first? Test the API for free in our sandbox environment. The sandbox version covers all basic signing. Through the use of sample data, you can see exactly how the API would function in your application.