Tutorial Part 1 CORPUS PROCESSING Kristina Kocijan krkocijanffzg
- Slides: 20
Tutorial Part 1 CORPUS PROCESSING Kristina Kocijan krkocijan@ffzg. hr Department of Information and Communication Sciences Faculty of Humanities and Social Sciences University of Zagreb
OUTLINE Text. not 1. � Create a new text � Import a text in any file format (ex. pdf / xml) � Open text file Corpus. noc 2. � Create a corpus � Open corpus Query 3. � Noo. J Reg. Ex Concordances 4. � Build concordances � Export concordances 5. 21. 5. 2012. Statistical analysis LREC 2012 - Noo. J Tutorial: Corpus Processing 2
TEXT UNITS = TUS *. not LREC 2012 - Noo. J Tutorial: Corpus Processing areas inside which linguistic resources are applied. 21. 5. 2012. Independant 3
TEXT: OPEN 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing 4
TEXT: IMPORT - PDF 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing 5
TEXT: IMPORT - XML 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing 6
TEXT: IMPORT - XML 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing 7
TEXT: CREATE NEW 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing Type new text. Copy -> paste text from another document. 8
CORPUS *. noc LREC 2012 - Noo. J Tutorial: Corpus Processing of text files that share the same linguistic resources. 21. 5. 2012. Collection 9
CORPUS: CONSTRUCTING 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing 10
CORPUS: OPEN EXISTING 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing 12
CORPUS: NEW 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing 13
LOCATE PATTERN – 4 TYPES OF PATTERNS 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing 14
LOCATE PATTERN – 3. NOOJ REGEX 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing lady young lady AND concatenation (a | the) lady parenthesis lady | girl OR disjunction <lady> <WF><lady> <lady>|<girl> <N+s+Hum> 17
EXAMPLES: WRITE A NOOJ REGEX (Mr. |Mrs. |Miss)<CAP> that will find all the words written in upper cases followed by any string of digits <UPP>(<D>)* If <E> = empty string write a Noo. J Reg. Ex that will find all the examples where ‘is’ is followed by 0, 1 or 2 any words that are followed by ‘the, this or that’ is (<E> | <WF>) (the|this|that) instead of ‘is’ recognize any form of verb ‘to be’ <be> (<E> | <WF>) (the|this|that) LREC 2012 - Noo. J Tutorial: Corpus Processing that will find all the Mr, Mrs and Miss followed by a. Name 21. 5. 2012. 18 between ‘to be’ forms and ‘the, this, that’ there can be any number of word forms <be> <WF>* (the|this|that)
LOCATE PATTERN? Will probably see Shall probably never see Is probably going to see Are probably about to see 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing 21
STATISTICAL ANALYSIS 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing 22
STATISTICAL ANALYSIS 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing 23
STATISTICAL ANALYSIS 21. 5. 2012. LREC 2012 - Noo. J Tutorial: Corpus Processing 24
Next Tutorial Part 2 LINGUISTIC UNITS AND ANNOTATIONS Max Silberztein max. silberztein@gmail. com University of Besançon
