DÉFINITION
An "affect" is a word or a group of words identified in a message. For each word or group of words detected, a tone is attached: positive or negative.
En quoi cela consiste-t-il ?
Affects are detected automatically thanks to machine learning technology. The Algorithm is based on a large organized dataset annotated manually. The tool will collect the expressions with the highest occurrence.
The tone is made thanks to the sentiment classifier (positive or negative) based on the expressions found within post contents.
Where to find them on Radarly ?
In your filter bar, Textual & thematic analysis section > Affects.
Then depending on your goals, you can select the positive or negative affects :
Note:
Affects can also be viewed in Posts & Analytics page by creating a personalized view in the workspace settings (need to be an administrator).
What are the benefits ?
Understanding of opinions expressed about a brand, product or a subject and more precisely how the authors qualify them: positive or negative.
Tag cloud allows you to quickly understand which expressions are most used in publications. The larger the size of the affect, the higher the occurrence.
This feature allows to get information about opinions in long text documents.
💡 Conseil
Vous avez encore besoin d'aide ? N'hésitez pas à nous contacter via le chat en direct ou à vous rendre sur la page de notre communauté de clients.
Trouvez des réponses et obtenez de l'aide de la part de l'équipe d'assistance de Meltwater et des experts de notre communauté.