About EmotionsRadar

EmotionsRadar is a research project and the name of the corresponding current software.

EmotionsRadar analyses emotions in texts (see “Demonstration of Results”). It can observe the development of emotions, e.g. anger (the shitstorm emotion) and has been running several Months in 2017 as a test version in the field of automotive. The evaluations were visualized in the frontend of a social media agency. Until end of 2017 the software is being optimized and tested.

The automated identification of emotions is also realizable for brands and products, organizations and stars in further industries and other languages.

EmotionsRadar is an opinon-mining-Software and based on many years of emotions research. Accordingly, it is not a simple off the shelf product, but still a research project - operated in cooperation with the Fraunhofer Institute.

Vision

One day, EmotionsRadar shall be able to show emotional profiles and moods to any search term at the push of a button.

Then users can enter a search word in the EmotionsRadar-Search Bar and the emotional profile of the term will be displayed.

The corresponding EmotionsRadar search bar could look like this:

Screenshot of the Search Bar
Screenshot of the Search Bar

Reason Why

Day-to-day, humans express millions of feelings in the Wold Wide Web. As emotions regulate the behaviour of humans, it is necessary to monitor them in the Web.

More than 300 emotional states exist, which can be expressed in written language. Moreover, the set of relevant emotions varies according to language, industry and brand. Consequently, the ability to identify and distinguish single emotions in big datasets makes EmotionsRadar unique.

Note: We analyse data. We do not generate, collect or sell them. Instead, we cooperate with social media agencies and extend their offer to the emotions analysis. Industry partners or forum operators also have the opportunity to provide us  their data for analysis, e.g. from their fora.

Set of Emotions

The identification of category-specific emotions is the basic requirement for any reasonable analyses. The following ten emotions cover 95% of the emotions found in German-language automotive fora:

LOVE

English
Sentiment: positive
Category: social
ICON: heart
--> This is a Love Story!

Emotional states connected with Love:

affection, commitment, confidence, pleasure, beeing in love, intellected, similarities, sympathy, feeling of safety, acceptance etc.

FRUSTRATION

  • Sentiment: negative
  • Category: social
  • ICON: broken heart

-->Reconcile!

Emotional states connected with Frustration:

antipathy, dislike (social), disesteem, envy, jealousy, disgust, feel rejected, feel alone etc.

JOY

  • Sentiment: positive
  • Category: situative
  • ICON: Smiley

--> Congratulations: Happy Customers!

Emotional states connected with Joy:

happiness, contentment, satisfaction, cheeriness, humour, fun, enjoyment, gag/joke, feel balanced etc.

SORROW

  • Category: situative
  • Sentiment: negative
  • ICON: Smiley gentle red

--> Look closer!

Emotional states connected with Sorrow:

concern, grief, pain, worrying about o.s./others / objects, suffering, sadness, scepticism etc.

ANGER

  • Sentiment: positive
  • category: situative
  • ICON: Smiley red

--> Warning: Angry Customers!

Emotional states connected with Anger:

frustration (situational), overreaching, to be under compulsion, defamation, suffering barriers, fury etc.

LIKE

  • Sentiment: positive
  • Category: cognitive
  • ICON: Thumbs up

--> They like it!

Emotional states connected with Like:

aesthetic senses: to take pleasure in sith, to enjoy / like objects / proportions / figures / the combination of colours / odours, intellectual understanding etc.

DISLIKE

  • Sentiment: negative
  • Category: cognitive
  • ICON: Thumbs down

--> Improve!

Emotional states connected with Dislike:

the opposite of the emotional states connected with "Like"

PRAISE

  • Category: cognitive
  • Sentiment: positive
  • ICON:Biceps

--> People belief in what you say/ do!

Emotional states connected with Praise:

"religious feelings", strongly believe in s.th., hope, exoneration, trust (cognitive), confidence, ransom etc.

REVIEW POSITIVE

  • Sentiment: positive
  • Category: cognitive
  • ICON: Ohr

--> Great! Customers are talking positively about your product

Emotional states connected with Review Positiv:

attention, interest, engagement, reference, purpose fuless, dedication (cognitive), judgement, positive findings/ learnings

REVIEW NEGATIVE

  • Sentiment: negative
  • Category: cognitive
  • ICON: Ear

--> Listen! Customers are complaining about the product mix.

Emotional states connected with Review Negative:

the opposite of the emotional states connected with "Review Positive"

Demonstration of Results

Below, examples of EmotionsRadar results and graphs are shown.

Note: The figures do no require to be representative or up-to-date for the brands mentioned there!

Level 1 - Shares of Emotions
Level 1 - Shares of Emotions
Level 2 - Emotions-Timeline
Level 2 - Emotions-Timeline
Level 3 - Emotions-Tagcloud
Level 3 - Emotions-Tagcloud
Performance in Automotive Dashboard 2017
Performance in Automotive Dashboard 2017
Quality of Statistical Models
Quality of Statistical Models
Performance Portfolio
Performance Portfolio

Key Figures

The 10 emotions that our EmotionsRadar software already recognizes in German automotive texts are explained below:

Key Figures Explanation Availability
OpinionAny emotionally coloured statement is an opinion and statements without emotional colouring are not opinions for us. EmotionsRadar ignores statements without emotion. It therefore does not provide any data noise, but focuses on emotions/opinions. That´s why the number of postings in our analyses is significantly lower than in conventional hits-based analyses. Only 2-8% of the texts in automotive forums contain emotions.YES
Number of Emotions

Counting the defined emotions (see Set of Emotions) found in texts. In one posting several emotions can be found.

YES
Share of Emotion

Shows the respective shares of the found individual emotions in texts to entered term, e.g. the Share of Joy. Out of 100% emotional expressions it shows the shares of the different emotions. The Share of Emotions can therefore be used to analyse the own composition of the 100% and to carry out a benchmarking with the competition (see Emotions Benchmark). The emotional profile of a brand should ideally match the brand´s emotional positioning.

See Level 1 - Shares of Emotions.

YES
Emotions-Tagcloud

Any emotion in the emotions benchmark can be analysed in depth by clicking on the emotion. When you select `Anger Audi´ for example, it will show you the terms that often appear in the context of `trouble with Audi´.

See Level 3 - EmotionsTagcloud.

YES
Emotions-Timeline

Shows the emotional shares for a term over time so that developments become clearly visible. For example changes in the emotional profile during campaigns or before/after launches, relaunches etc.

See Level 2 - Emotions-Timeline.

YES
Techterms-FilterFor our automotive analyses a so called Techterms filter can be activated for the Emotions Tagcloud. It shows terms that are regularly relevant to cars if you activate it. For example technical parts (such as engine) or terms related to the car dealer or car workshop.YES
Perfomance PortfolioThis portfolio shows the emotional positioning of different brands in direct comparison to a selected emotion. For example the “Share of Joy” (in %) of several brands can be displayed on the x axis and the shares of opinions on the y axis.NO

Milestones

  1. Start of Basic Research

    2002

    Start of Basic Research for EmotionsRadar on Emotions and Choice of Brands by Nicole Krautkrämer-Merkt (PhD Thesis).

  2. First Offline Emotions Studies

    2004

    First offline emotions studies are commissioned by customers after the German special interest magazine „Absatzwirtschaft“ published the first research results.

    Article in Absatzwirtschaft „Wenn Gefühle in die Irre führen“

  3. Completion of Doctoral Thesis

    2006

    Completion of the dissertation by Nicole Krautkrämer-Merkt

  4. Start Fraunhofer Cooperation

    2007

    Start of the collaberation with Fraunhofer Institute, with the common goal of identifying emotions in online texts.

    Article in weiter.vorn, Fraunhofer Magazin, Beilage Theseus Seite 24 „EmotionsRadar erkennt Stimmungen“

  5. First CeBit Hannover

    2009

    Presentation of the first monitoring results of a new car launch generates a very positive response from the media and trade press.

    Article in Absatzwirtschaft „Vom Fachbeitrag zum Tool“

    Article in heise online „EmotionsRadar erkennt Gefühle in Online-Foren

    Article in the Rheinpfalz „Schürfer im Datenbergwerk“

  6. Tests & Feedbacks

    2010

    Additional tests and discussions with the industry and policy about concept idea and results.

    Article in Planung & Analyse „Was Wähler im Internet bewegt“

  7. Foundation of EmotionsRadar GmbH

    2011

    The informal cooperation of Krautkrämer-Merkt and Fraunhofer researchers, which has hitherto been pursued through research interests, has now been formally and contractually regulated. In consequence: Founding EmotionsRadar GmbH

  8. Second CeBit Hannover

    2013

    EmotionsRadar is presented as an example of a `BigData Machine.

  9. dmexco Cologne

    2014

    Demonstration of a first demo and feedback discussions with experts, a.o. motor-talk.de – Europes largest car and motor community.

  10. Admission of two Investors

    2014

    Admission of two investors/business angels to company. Goal: To enter the market of Social Media Monitoring for Automotive.

  11. Start Hardware R&D Project

    2016

    The Hardware requirements of the Big Data Software become clear. Start of a separate research project on this topic at University of Applied Sciences in Koblenz.

  12. Test Automotive Sector

    2017

    Basic EmotionsRadar function were for the first time and for testing installed the Systems of a Social Media Agency (especially for R&D purposes).

    The test is finished and currently the software is being further optimised.

    see E-Commerce Magazine Sommer 2017

  13. Update/ Further Optimization EmotionsRadar Software

    2017

    Internal Test of the updated Version of EmotionsRadar are going on.

Contact

Please do not hesitate to contact us.

Your questions about EmotionsRadar and further research and projects answers

Rebecca Fröhlich, Project Manager

+49 (0) 261 97 39 788 -1

Use our contact form or send an e-maill to:

Prof. Dr. Nicole Krautkrämer-Merkt
Geschäftsführende Gesellschafterin
und Gründerin der EmotionsRadar GmbH

n.merkt@emotionsradar.com

This website about EmotionsRadar is owned by:

EmotionsRadar GmbH
Im Vogelsang 8
56179 Vallendar
Germany

HRB 22607, Koblenz

Steuer-Nr. 22/653/03537

Contactinformation
What do you want to tell us?

* required fields