Skip to main content

Service classification - STILINGUE Smart Care

  • Updated

Index:

  1. How it works?
  2. How it works?
  3. How to activate classification?
  4. Categories

 

Discover the possibility of automatic service classification in STILINGUE Smart Care, using our Artificial Intelligence engine.

Those who work in customer service using social media know that efficiency and speed are essential for analyzing, classifying, and responding to interactions. With this in mind, STILINGUE based its approach on the concept of Voice of the Customer (VoC), which relates to the process of understanding customer expectations, demands, and pain points to bring more quality to automatic classification by the main service themes in your conversations through the platform. 

 

How it works?

STILINGUE conducted various studies by selecting some of the largest customer service operations in Brazil to identify the main categories used daily by these companies. This allowed us to understand how experts practice the segmentation of services within each operation and how crucial it is to save time. After gathering the necessary information, our Artificial Intelligence (AI) team created a classifier using Machine Learning - a branch of the AI field - to automate the classification of services within the most commonly practiced categories found in the market. 


How it works?

The customer service classifier is capable of understanding Portuguese (BR) and automatically identifying the category to which the interaction belongs. This means that it organizes what has been collected by STILINGUE Smart Care and automatically applies the pre-established categories on the platform. This results in faster and higher-quality analysis of service requests. Moreover, it provides a real-time overview of the types of interactions consumers are engaging in.

Below are the pre-defined categories for automatic classification, used by major customer service operations in Brazil:

  1. Gratitude;
  2. Question;
  3. Compliment;
  4. Recommendation;
  5. Interest;
  6. Suggestion;
  7. Complaint.


How to activate classification?

If you are interested in using the STILINGUE Smart Care automatic classification service for your customer service operation, you should get in touch with the Customer Service team through the tool's chat. 

Once the functionality is internally released by our team, you will be informed about the possibility of choosing the quantity of categories (you don't need to use all of them) and whether their nomenclature aligns with your operation. 

These details will be communicated to the internal team, and after that, you can register the terms in the Search Configuration. Create a theme for each desired category (as described above), without adding any terms within the fields.

Note: It is also possible to create categories as tags within the Search Configuration.

Here's how to locate the Edit Search area and the related fields:

 

Step 1:

 

Step 2:

 

Afterward, the classifier activation is done internally by the STILINGUE team. You will be notified when interactions collected in STILINGUE Smart Care (belonging to each category type) will be automatically classified using the Machine Learning model. 

Conversations classified automatically by the classifier will be identified with a gear symbol in front, distinguishing them from the rest of the system's standard classifications.

 


Categories

To better understand the classification, it is important to know what each category encompasses:

1. Gratitude 

Any text construction that contains words related to "thank" concerning the use or functionality of products or services from the company is considered gratitude. These constructions are also characterized by short sentences. 

Keywords: thank you, thanks, thanks a lot, cheers, thank you so much, thx, thanks a bunch, thanks a million, thanks a ton, thanks a bunch. 

2. Question 

Questions are characterized when the user in question expresses any uncertainty about functionality or service. It is also characterized by the use of adverbs and question marks.

Keywords: how do I, how can I, how does it work, what's happening, how do I use, doubt; what's going on? why aren't you answering me? why…? can you explain...; I wanted to know...; where do I go? is it.

3. Compliment 

Compliment is characterized by the user's satisfaction with the service provided by the company. Positive adjectives and words of contentment are also observed in this criterion. 

Keywords: admiration, admire, love, adore, adored, beloved, loved, we love, loved, love very much, love.

4. Recommendation 

Internet users often recommend services and products to people in their contact network. In this context, recommendations stand out through tagging individuals and using indicative words. 

Keywords: there it is, as I told you, look at this, it's this one.

5. Interest

Text construction that involves any type of interest, desire, or intention to purchase related to the company's products and services.

Keywords: I want, we want, treats, let's go, I'm interested, interest, I will buy, I want to buy, I want a limit, increase my limit.

6. Suggestion 

Semantic features related to suggestions stand out through suggestive expressions, as described in the keywords.

Keywords: a tip, a suggestion, it would be nice, it would be better, could have, if it had, how about, it would be good, think carefully, think carefully, just missing this.

7. Complaint 

By complaint, we mean text construction expressing dissatisfaction with the services offered by the company. Negative adjectives and words of discontent are some of the markers used by users. 

Keywords: disappointment, disappointed, disappointing, consumer defense, leaves much to be desired.

 

For more information, visit the discussion on the subject at our community or videos on our channel. 😃

Related articles