Lexical Affect Sensing: Are Affect Dictionaries Necessary to Analyze Affect

1
Lexical Affect Sensing Are Affect Dictionaries
Necessary to Analyze Affect?

• Alexander Osherenko, Elisabeth André
• University of Augsburg

2
What emotions conveythese textual utterances
(SAL corpus)?

• High arousal, negative valenceNo, well, I'm not
  a fool.
• High arousal, positive valenceNo, ltlaughgt I
  think I'm being stupid actually.
• Low arousal, positive valenceYup.

3
Dictionaries

• Dictionary of Affect Language (DAL - Whissell)
• happy (evaluation, activation, imagery)3.0000
  2.7500 2.2
• Linguistic Inquiry and Word Count Dictionary
  (LIWC)
• happy (categories)Affect, Positive emotion,
  Positive feeling
• BNC frequency list11649 happy aj0
• SAL frequency list

4
Research questions

• Are recognition rates higher if word features are
  emotional?
• Do emotive annotations in affect dictionaries
  improve recognition?
• Are common words more useful than less common
  words?
• Are dictionaries of affect more useful than
  general-purpose dictionaries?

5
Feature Extraction and Evaluation

• Word features
• Selection of the most expressive words
• Selection of the most frequent features
• LIWC features (CAT-68 and CAT-8)
• DAL features (EA-AVG)
• Average values for the evaluation, activation,
  imagery scores

6
Evaluation

• 672 utterances from the SAL corpus as a
  5-classes-problem
• The majority vote strategy
• The SVM classifier
• Averaged recall value/number of word features

7

• Useful criterion of feature reduction without
  risking a severe degradation of recognition rates

8
Do emotive annotations in affect dictionaries
improve recognition?
Affect-related features do not include
discriminative information that is not yet
included in the word counts
9
Hard to say whether a reduction of features
should be based rather on the frequency of words
or their expressive qualities
10
General-purpose dictionaries may provide similar
results as affect dictionaries for similar
numbers of features
11
Recommendations

• Frequency strategy is not worse than the
  emotional expressivity strategy
• Similar trends for a movie reviews corpus
• Results dont degrade dramatically when reducting
  number of word features (real-time recognition)
• Acceptable results also with only affect
  annotations

12
Thank you!
13
Conclusion
14
Mapping of FEELTRACE data onto affect segments
1.0
Activation
high_neg
high_pos
neutral
0.2
Evaluation
1.0
low_neg
low_pos

• Examples
• Affect segment high_pos (Laugh) I'm damn
  awful. How are you (laugh)?
• Affect segment low_neg Erm, that's probably
  true.