Analysis%20of%20sentiment%20syntagma%20using%20dependency%20tree

1
Analysis of sentiment syntagma using dependency
tree

• Serge B. Potemkin
• Moscow State University
• potemkin_at_philol.msu.ru

2
Terms

• Sentiment
• A thought, view, or attitude, especially one
  based mainly on emotion instead of reason
• Sentiment Analysis (opinion mining)
• use of natural language processing (NLP) and
  computational techniques for extraction or
  classification of sentiment from (unstructured)
  text

3
What for?

• Consumer information
• Product reviews
• Consumer attitudes
• Trends
• Politics
• Politicians want to know voters views
• Voters want to know policitians intentions and
  who else supports them
• Social
• Find like-minded individuals or communities
• Financial
• Predict market trends given the current
  opinions

4
Features

• Which features to use?
• Words (unigrams)
• Phrases/n-grams
• Sentences
• How to interpret features for sentiment
  detection?
• Bag of words
• Annotated lexicons (WordNet, SentiWordNet)
• Syntactic patterns
• Paragraph structure

5
Challenges

• Harder than topical classification, with which
  bag of words features perform well
• Must consider other features due to
• Ambiguity of sentiment expression
• irony
• expression of sentiment using neutral words
• many others
• Domain/context dependence
• words/phrases can mean different things in
  different contexts and domains
• Effect of syntax on semantics

6
Formal description

• Semantic orientation of a sentence expressed by a
  ternary predicate
• O(subject, object, sentiment)
• sentiment bad, neutral, good
• i.e.,
• the subject of assessment considers the object
  of assessment to be good or bad (or neutral
  not a sentiment)

7
Sentiment expression in NL

• Predicate O may be expressed explicitly (Vania
  likes Masha) -
• only the surface syntactic analysis is needed
• Vania (subj) likes (sentiment) Masha (obj)
• to determine its semantic orientation (SO).
• The common case is quite different(Vania
  suffers from Mashas absence) both suffer and
  absence are negative but the sense is equivalent.

8
Bag of words vs. syntagma an.

• Bag of words (number of positive and negative
  words) gives good results for large texts
• Syntagma a phrase forming a syntactic unit,
  say modifier (X) keyword (Y)i.e.
  adjectivenoun or adverbverb
• Signature of syntagma SO sgn(X,Y,neg/0/pos).

9
SO Calculus

• ? X,Y.sgn(X,Y,pos) ? dep(mod,X,Y),sgn(X,pos),sgn(
  Y,pos). (a)
• i.e. if X,Y positive then XY positive
• ? X,Y,Z.sgn(X,Y,Z) ? dep(mod,X,Y),sgn(X,0),sgn(Y,
  Z). (b)
• i.e. if X pos., Y neut. then XY pos.
• ? X,Y,Z.sgn(X,Y,Z) ? dep(mod,X,Y),sgn(X,Z),sgn(Y,
  0). (c)

10
Different orientation of syntagma constituent
words

• sgn(????????,???????,pos)sgn(mad,happyness,pos),
  sgn(???????,?????,pos)sgn(furious,success,pos)
  ,
• sgn(????????,?????,neg)sgn(considerable,damage,n
  eg), sgn(???????,???????,neg)sgn(good,scolding,
  neg).
• Kustova, 1

11
Ambigoues cases

• sgn(?????,???,?), sgn(??????,?????,?)
• sgn (bad,peace,?), sgn (good,war,?)
• The expression "a bad peace is better than a good
  war," establishes an order relation "better"
  among its member attributive constructions, but
  one can assume that both are bad, i.e., sgn
  sgn(bad,peace,neg), sgn(good,war,neg). In some
  other context, "good war" could be perceived as a
  positive phenomenon.

12
Double negative

• Logical rule of double negation
• ? X,Y,Z.sgn(X,Y,pos) ? dep(mod,X,Y),sgn(X,neg),
  sgn(Y,neg).
• fails in NL
• weak opponent, impotent aggressor, toothless
  criticism (neut.)
• or
• bitter sorrow, blatant outrage, brutal torture
  (neg.)

13
Syntagma evaluation

• Methods
• expert evaluations performed by several
  independent experts Osgood,2, who are asked to
  mark up SO of isolated words and syntagma,
  assigning them a label pos/0/neg
• corpus techniques, performed on an
  sentiment-annotated corpus Zagibalov,3,
• SentiWordNet

14
SentiWordNet

• Based on WordNet synsets
• http//wordnet.princeton.edu/
• Ternary classifier
• Positive, negative, and neutral scores for each
  synset
• Provides means of gauging sentiment for a text

15
SentiWordNet Construction

• Created training sets of synsets, Lp and Ln
• Start with small number of synsets with
  fundamentally positive or negative semantics,
  e.g., nice and nasty
• Use WordNet relations, e.g., direct antonymy,
  similarity, derived-from, to expand Lp and Ln
  over K iterations
• Lo (objective) is set of synsets not in Lp or Ln
• Trained classifiers on training set
• Rocchio and SVM
• Use four values of K to create eight classifiers
  with different precision/recall characteristics
• As K increases, P decreases and R increases

16
SentiWordNet Results

• 24.6 synsets with Objectivelt1.0
• Many terms are classified with some degree of
  subjectivity
• 10.45 with Objectivelt0.5
• 0.56 with Objectivelt0.125
• Only a few terms are classified as definitively
  subjective
• Difficult (if not impossible) to accurately
  assess performance

17
Corpus-based method

• Sentiment annotated corpora (English and Russian)
  of approx. 1500 short utterances concerning
  popular books. Each utterance contains from 1 to
  15 sentences and was marked with a label neg /
  pos.

18
Corpus processing

• - Stemming and determination of morphological
  characters of each word (without morphology
  disambiguation)
• - Parse with obtaining the dependency tree for
  each sentence Potemkin, 4
• - Joining the particle "no/not" to the associated
  word (not understand gt not_understand)
• - Selection of constructions modifierkey word
  (adjectivenoun, adverbverb)
• - Counting the number of occurrences for each key
  word nverb,

19
Corpus processing (continued)

• - Counting the number of occurrences in the
  positive-marked utterances nvp and negatively
  labeled utterances nvn
• - Calculation of the normalized assessment factor
  for each key word kv (nvp-nvn) / nverb
• - The same calculations for each modifier to give
  the normalized assessment factor kd, and for each
  syntagma in the corpus - the normalized
  assessment factor ks.

20
Assessment thresholds

• Assessment factors ks ? -1,1,
• ks ? -1, -0.6) neg ks ? -0.6, 0.6 0
  ks ? (0.6, 1 pos

21
Table of syntagma signatures
neg -key 0 -key pos -key
neg -mod neg not_palatable demagogy pos defeated enemy neg uninteresting book pos forgotten kingdoms neg banal action-film pos secondery pleasure
0 -mod neg star fever pos imminent defeat neg unexpected level. pos only book. neg. late success pos continues growth
pos -mod neg happy end pos fine rubbish neg good intentions pos pleasant book neg sweet honey pos best masterpiece
22
Histogram of syntagma distribution over the texts
23
Histogram of the 1st word of syntagma
distribution
24
Histogram of the 2nd word of syntagma
distribution
25
Conclusion

• The report presents considerations for
  determining the sentiment of syntagma on the
  basis of evaluation of the signature of its
  constituent words for structures such as
  adjectivenoun, verbadverb.
• Logical formulas specifying the calculation of
  semantic orientations are listed.
• An experiment over the semantically annotated
  sentences was performed.
• The further research concerning predictive
  syntagma of type subject verb object will be
  undertaken.

26
References

• http//dict.ruslang.ru/magn.php
• Charles E. Osgood, George Suci, Percy
  Tannenbaum, The Measurement of Meaning.
  University of Illinois Press, 1957.
• http//www.informatics.sussex.ac.uk/users/tz21/
• http//sunsite.informatik.rwth-aachen.de/Publicati
  ons/CEUR-WS/Vol-476/paper6.pdf