Language

Let us introduce you to one of the features (not only) international teams love! ❤️ Yes, the "Language" module with some pretty neat components like translating, analysing the sentiment, or detecting the language is what we have in mind.

We worked pretty hard on this one to empower you with some cool language-based features. As outlined earlier, there are three main components under the "Language" module:

• Analyse the sentiment

• Detect the language

• Translate

Let's look at an example using the "Analyse the sentiment" component. Today, brands don't fully control the way they are perceived. More than ever, (potential) customers shape the way our brand is sensed.

Having said that, let's imagine we want to listen for mentions on Twitter (when someone mentions our brand) and analyse the sentiment of the message. If the calculated score of the sentiment is lower than a certain limit (zero in this case), we want to get notified on Slack. This way, we make sure no Tweet that puts our brand in a bad light gets forgotten and unanswered. To celebrate positive brand mentions with the team, let's send a Slack message every time a positive mention is Tweeted.

Below, see what the complete flow looks like in Appmixer.

Preview of a flow using the "AnalyseSentiment" component

Makes sense? Great, let's uncover the other two components: detect the language and translate.

Say your customer support gets requests in many languages of which some are not supported. Imagine you'd want to detect the language of the ticket and send it to a Google Sheet for statistical purposes. Below, see such an example using Freshdesk component that listens for new tickets.

Detecting the language of a customer request and passing it to Google Sheet

As a result, we will be able to detect the language and make decisions based on the outcome. Now, let's extend the current flow and translate the request to English using the third component: translate. Besides that, we may want to update the ticket using the translated text to keep things organised for our customer support.

Translating the ticket and updating it in Freshdesk

God damn, that feels good! 😎We just eased our customer support of some boring manual work.

Before we call it a day, let's answer some of the questions related to the "Language" module.

Frequently asked questions

Can I analyse the sentiment of a plain text and HTML? Yes, you can! We support both plain text and HTML.

How is the "Score" of the sentiment calculated? The score indicates positive sentiment with a value greater than zero, and negative sentiment with a value less than zero. Refer to the Google documentation to learn more.

What does the "Magnitude" in the AnalyseSentiment component mean? The magnitude of a document's sentiment indicates how much emotional content is present within the document, and this value is often proportional to the length of the document. Refer to the Google documentation to learn more.

What engine stands behind the text translation? We use Google's Cloud Translation API which has been improved dramatically over the past years. You can, nonetheless, build your own engine and incorporate in to your Appmixer instance.

What languages are supported for the translation? Currently, more than 130 languages are supported. You can find the definitive list in the Google documentation: https://cloud.google.com/translate/docs/languages

How do you calculate the "Confidence" used in the DetectLanguage component? Confidence is a range from 0 to 1. 1 is 100% confident. Learn more about this feature here.

In the next section, we'll learn how accounts work in Appmixer.