Annotation Guidelines for the Penn Discourse Treebank Part B

1
Annotation Guidelines for the Penn Discourse
TreebankPart B

• Eleni Miltsakaki, Rashmi Prasad, Aravind Joshi,
  Bonnie Webber

2
Brief summary

• Annotation of discourse connectives and their
  arguments.
• Discourse connectives subordinate conjunctions,
  coordinate conjunctions, adverbials, empty.
• Discourse connectives express relations between
  at least 2 events or states.
• Legal argument a clause at minimum.
• Annotation tool WordFreak
• Tags CONN,ARG1, ARG2, SUP1, SUP2
• Features Search, discontinuous text selection,
  comment box.

3
Basic types of clauses

• Tensed
  Non-tensed
• Main Subordinate
  Subordinate
• Complement Adverbial Relative Infinitival
  Participial

4
Examples

• Main Tom left
• Complement Mary said that Tom left
• Adverbial Tom left when he finished
• Relative Tom, who finished early, left
• Infinitival Tom wants to leave
• Participial Tom spend the day watching TV

5
Small clauses

• Complements (objects) of certain verbs
• Verb be is understood but may not be explicit
• (1) I consider Mary a smart student.

6
Small clauses as arguments

• Selecting just the small clause is sufficient
  even though there is no explicit verb
• (2) I consider Mary a smart student
  although she failed her exams.

7
Relative clauses as arguments

• Selecting just the relative clause is sufficient
• Syntactic information from Treebank will help us
  identify the head of the relative clause
• (3) I bought some booksi which (nulli) were very
  expensive, even though they were second hand.

8
Distinction between a relative clause and an NP

• a reference to spiders that attract males and
  then kill them after mating.
• that attract males and then kill them after
  mating ? relative clause ? OK ARG
• spiders that attract makes and then kill them
  after mating ? Noun Phrase ? NOT OK ARG

9
Modified connectives

• As other syntactic categories, connectives can be
  modified.
• E.g., only when, largely because, especially
  after, etc.
• In such cases, select both the connective and the
  modifier.
• When you see a comma before the connective,
  select just the connective. In such cases, the
  modifier does not modify the connective.
• (4) He wears jeans only, because he wants to
  have a casual look.
• (5) He wears jeans only because he wants to
  have a casual look.

10
Words that look like discourse connectives

• Reminder
• Discourse relations require clausal
  interpretations.
• Ignore instances of connectives in your set if
  they are not associated with a clause
• Examples
• (6) These mainly involved such areas as materials
  --advanced soldering machines, for example and
  medical devices derived from experimentation in
  space.
• (7) They bought wine and beer.
• (8) Mary, also John, will leave late today.

11
Not all adverbials are connectives

• Some adverbials do not express a discourse
  relation. The clause that contain them is
  sufficient for the interpretation.
• An adverbial counts as a connectives when it
  expresses a relation between at least TWO
  situations in the discourse.
• In
• (9) John did not finish the report.
  Therefore, we will postpone todays meeting.
• Out
• (10) John was hungry. Strangely, he only
  ordered a fruit salad.

12
ARG1 and ARG2 for double connectives

• For double connectives such as On one hand on
  the other hand, Ifthen
• Select the two connectives using the
  discontinuous text selection feature and enter
  them together under CONN.
• Mark as ARG1 the clause that contains the first
  connective.
• Mark as ARG2 the clause that contains the second
  connectives.
• (11) If you finish your homework before noon,
  then you may go to the movies.
• ARG1 you finish your homework before noon
• ARG2you may go to the movies

13
A few more conventions

• Exclude punctuation marks appearing at the end of
  the clause that you are selecting.
• When selecting the content of a complement
  clause, include that in your selection.
• When selecting a relative clause, include the
  relative pronoun in your selection.
• When a connective appears in the clause youre
  selecting as an argument, include that connective.

14
The case of VP coordination

• Do not annotated connectives that related two
  verb phrases.
• Diagnostic missing subject in the second
  (tensed) verb.
• (12) OUT Mary finished her food and left.
• (13) IN Mary finished her food and she left.

15
Some hard cases As

• Multiple meanings of as
• Temporal
• (14) He tripped over a bunch of plugs as he was
  leaving the room.
• Causal
• (15) W. U. had major losses as its telex
  business faltered in the face of competition from
  facsimile machines.
• Annotate as only if it has a temporal or causal
  interpretation.
• Do NOT annotate, for example
• (16) As she puts it, theres no hope.
• (17) We do as we are told, as is the rule.

16
Some hard cases So

• So expresses a consequence relation.
• But its not always easy to identify the
  consequence
• (18) She flunked the exam. So, whats next?
• (19) You said she didnt work hard. So, if you
  believe this, you must be right.
• Sometimes it may be hard to identify ARG.
• If you run into such cases, let us know and make
  a comment in the comment box.

17
Some hard cases Nor

• Nor can be found
• By itself annotate as regular connective.
• (20) This has nothing to do with you. Nor will it
  ever.
• In a neither nor construction annotate as a
  double connective.
• Sometimes it may be hard to identify ARG1. In
  this case, leave the ARG1 slot empty and make a
  comment in the comment box.
• Neither nor in VP coordination Do not
  annotate
• (21) In doing so, he neither rejected a socialist
  planned economy nor embraced the free market.

18
When to exclude conns from args

• Do not include a connective in the selection of
  an argument if it does not belong with the clause
  selected as an argument.
• Some such hard cases include the coordinate
  conjunctions and and but.
• But make sure you include the connective when it
  belongs with the selected clause even if its at
  a distance.
• (22) But, say Mr. Dinkins, he did get an office.
  Therefore he shouldnt complain.

19
Implicit conns multiple interpretations

• If you identify more than one relation between
  adjacent sentences
• And therefore are able to provide more than one
  explicit connective
• Put the one that you think most likely first and
  add the rest as follows
• CONNbecause CONNnevertheless
  CONNmoreover
• If you think there is NO relation between the two
  sentence that can be expressed with a connective
  type NONE in the comment box. Do not just leave
  the comment box empty.