Annotation Guide

• Intro
  • Overview
  • What’s Covered?
  • See Also
• Setup
  • Xcode
    • Installing Xcode Tools on Mac OS X
  • Subversion
    • Creating directory for repos
    • Checking out the morphosyntax repo
    • Checking out the data repo
    • Checking out the code repo
    • Final steps for setting up repos
  • CLAN
    • Installing the GUI
    • Installing the command line tools
  • Edit your profile
  • Edit vimrc
• Repository
  • bin
  • lib
  • clan
    • clan/lib
    • clan/bin
  • chat
    • chat/incoming
    • chat/coders
    • chat/final
    • chat/inserts
• Programs
  • all
  • add
  • caps
  • commitlex
  • compare.py
    • Usage
    • Options
  • dic
    • Usage
  • dis
  • fixlines
  • grasper
  • postal
  • root_nv
  • synflagger
  • syntax_extract
  • Helper programs
    • dict
    • fixpost.pl
• The Annotation Process
  • 1. Navigate to your working directory
  • 2. Use dump.py to generate the CHAT file
  • 3. Clean up the file for morphosyntactic analysis
  • 4. Reconcile words not found in the CLAN lexicon
    • 4.1. Spelling errors
    • 4.2. Compound words
    • 4.3. Numbers
  • 4.4. Applying the & and @ symbols to non-words or idiomatic words
    • 4.5. Adding words to the CLAN lexicon
  • 5. Run MOR to generate a .mor.cex file
  • 6. Run the automatic part-of-speech disambiguator POST
  • 7. Manually disambiguate the remaining ambiguous words in the .pst.cex file
  • 8. Generate the syntax tier with GRASP
  • 9. Run fixlines on the newly created syntax file
  • 10. Flag potentially incorrectly coded utterances with graspParse.py
  • 11. Search for flags and correct any problems
  • 12. Commit the corrected syntax-coded transcript back into the SVN repository
• Prepping a transcript
  • 1. Remove disfluencies and repetitions
    • 1.1. Remove utterances consisting only of #
  • 2. Find and remove repeated words and phrases
    • 2.1. Use the tn shortcut in VI to find single-word repetitions
    • 2.2. Use the program dis to find repeated words and phrases
  • 3. Add an & symbol to the beginning of non-words
  • 4. Ensure that “Mom”, “Dad”, etc. are properly capitalized
    • 4.1. Find incorrectly lower-case family terms using the `m shortcut
    • 4.2. Find incorrectly capitalized family terms using the `d shortcut
  • 5. Search for and correct common mistakes using the `v shortcut
    • 5.1. Phrases that need or might need to be joined by a +
    • 5.2. Phrases that need or might need to be changed or written with an apostrophe
    • 5.3. Two capitalized words in a row
    • 5.4. Punctuation placed in the middle of an utterance
    • 5.5. Repetitions marked by [xN]
    • 5.6. Single quotes (apostrophes) used as quotation marks
    • 5.7. Apostrophes in proper nouns
    • 5.8. Misuse of the @l marker
    • 5.9. A space at the beginning of the tier
    • 5.10. Utterances missing punctuation
  • 6. Run caps to verify that capitalized words are proper nouns and are correctly formatted
• Common problems
  • Abbreviations
  • Part of speech
    • The phrase go to sleep
  • Syntax
    • V NP to V
      • XCOMP
      • XMOD
      • XJCT
    • V NP V
    • AUX and INF as head of a verb cluster
    • DET as the head of a noun phrase
    • det used as JCT
    • Nominals in apposition
    • Pronouns used as determiners
    • Missing copulae
    • Missing copula vs. post-positioned MOD
    • Post-positioned MODs after -thing, -body, and -one words
    • SUBJ after ROOT
    • Post-positioned quantifiers
    • Perfect participle as PRED vs. passive construction
      • As PRED
      • As passive construction
    • Compound (multi-word) prepositions
    • Prepositional phrase as object of another preposition
    • Dangling COORDs
    • When to use PRED
    • Using the TAG GR
      • Negated auxiliary/copula tags
      • Sentence fragment with full verb
      • Full sentence with no grammatical connection
    • The phrase to be + done/finished + VERBing
    • Quantifiers and quantifier + of phrases
    • When to use VOC
    • Structure of a noun phrase
    • COMPs and CPREDs introduce by WH-words
    • all NP VERB is
    • The phrase be ADJ to VERB
    • The phrase good at VERBing
    • Clauses used as objects of prepositions
    • WH-word do you think CLAUSE
• VI Shortcuts
  • General use shortcuts
  • Changing GR information
  • Changing part-of-speech information