Release Notes 15.12.2020

Find here an overview of the newest releases we shipped to the platform

👍New Features

  • Brand perspective on psychographics
    Psychographics can now be compared by meta fields (e.g. brand, location, etc.) much more conveniently with the latest update of the psychographics page. If you choose to compare by brands, you’ll see how the personality traits and communicate styles differ across the available brands. You’ll also learn which of the individual traits of which brand are over or under indexing compared to the average.


  • Reuse topic model
    This feature allows you to apply the structure of an existing project with all its customizations to a new project. When creating a new project, rather than selecting from the list of available use cases, you can select an existing project to apply its model to your new project. 


  • View and edit the polarity of a token
    You can now assign a polarity to a token if you expect the token to be always positive or negative when it appears.

    A prominent example is the word “expensive”. It is considered as a negative polarity token, because “expensive” is negative by its nature. However, if negated (e.g. not expensive), it becomes positive.

    Another scenario is the use of a sentiment word in relation to a polarity token. For example, the positive polarity token “accessible for wheelchairs”, will become negative if a negative sentiment word is detected close to it (e.g. poorly accessible for wheelchairs").