Interpreting a text is much more than understanding its content. When it comes to communication, the medium, the emotions and the context are also of great importance in fully understanding a message. These characteristics can be expressed in text. And this is where the data can be confusing.
Words can express different things. And it depends on the emotion of the author. Dissatisfaction, happiness, sadness, anger and trust are examples of emotions that can be easily interpreted by us humans during a reading. However, the current demands are too high to assign this task to employees. Then it is given to artificial intelligences These, if not well trained, can give inappropriate feedback by discarding the various characteristics contained in a message.
In order to bypass this problem, IBM has created a set of tools that interpret emotion, tone, and meaning present in every text or document. One of them is the Watson Tone Analyzer.
It works as follows: the user sends an entry in JSON, plain text or HTML (up to 128 KB of text). Following, Tone Analyzer returns the results in JSON, with the tone of the text already highlighted.
IBM offers a demo of the tool with several examples – including tweets – click here to check!
Contents
Understand better, for better answers
Online service is a dilemma. Designating people for this role is not as effective as using artificial intelligence. However, human-to-human communication is less plastered and limited. This provides greater customer comfort.
Recommended: What is the ideal connection speed to run Google Stadia well?
Watson’s tools solve this. Watson Tone Analyzer and Watson Discovery work together to understand the messages and respond appropriately.
Outside of corporate domains, Watson tools are able to evaluate social comment. And they provide accurate feedback on how customers react publicly to products.
The marketing of a company can also benefit from Tone Analyzer. The tool is able to evaluate a customer’s mood and customize the advertisement targeted to it. This is extremely useful for delivery applications, travel companies, streaming services. It could change their suggestions, destinations, terms or playlists for better adapt to the mood of the user.
Another context that can be improved through these tools is the customer service (SAC) using humans. With Watson, supervisors can observe the workings of employees and how satisfied customers are.
Curious Applications
The famous TED organization applied Watson’s tools to study audience reactions during TED Talks. A total of 1,931 presentations were searched for patterns that generated more applause.
Curiously, the audience’s positive reaction was directly related to the speaker’s posture. It’s omething that could be identified using the Tone Analyzer. The presenter who expressed the most sadness or impersonality would receive less applause. While the more confident, cheerful and satisfied would have a better chance of provoking more heated reactions.
Another quite unusual application is on dating sites. Analyzing a user’s temperament, relying on his writing, increases the platform’s chances of finding the ideal partner.
For this, the researchers applied the Tone Analyzer in 50 thousand registered profiles. They classified the conversations already made as “potential pairs”. And they compared the tones of the users to create correlations and, later, a statistical model. Finally, they compared this model to other baselines that consider other attributes.
The result? A strong correlation was found between similar tones and the number of messages exchanged. More precisely, this led to an improvement of 45% in the dialogs compared to the old ones.
To use the Watson tools, there are 3 ways. The first of them is signing up for free with IBM Cloud Lite. However, by this method there are some limitations. Alternatively, choose the IBM Cloud services you want. And with the help of the IBM calculator, determine the subscription price. Finally, the last option is to contact IBM and hire a fixed price plan. It’s a method for companies and development teams.
Tone Analyzer is currently able to interpret English and French texts.
Also Read: Millions of Facebook and Instagram passwords were exposed to employees
Leave a Reply