TEXT STATISTICS 3 DAY /24/14 LING 3820 & 6820 Natural Language Processing Harry Howard Tulane University

Course organization 24-Oct-2014NLP, Prof. Howard, Tulane University 2   The syllabus is under construction.   Chapter numbering  3.7. How to deal with non-English characters 3.7. How to deal with non-English characters  4.5. How to create a pattern with Unicode characters 4.5. How to create a pattern with Unicode characters  6. Control 6. Control

Open Spyder 24-Oct NLP, Prof. Howard, Tulane University

>>> from corpFunctions import textLoader >>> text = textLoader('Wub.txt') >>> from nltk.probability import FreqDist Review of dictionaries & FreqDist 24-Oct NLP, Prof. Howard, Tulane University

NLTK version  import nltk  print nltk.__version__ 24-Oct-2014NLP, Prof. Howard, Tulane University 5

Clarification of dictionary methods 1. >>> tallyDict = {'all':13, 'semantic':1} 2. >>> tallyDict['all'] 3. >>> tallyDict['pardon'] 4. >>> tallyDict['pardon'] = 1 5. >>> tallyDict['pardon'] = 2 6. >>> tallyDict = {'all':13, 'semantic':1, 'semantic':5} 7. >>> tallyDict[13] 24-Oct-2014NLP, Prof. Howard, Tulane University 6

A dicionary maps a key to a value 24-Oct-2014NLP, Prof. Howard, Tulane University 7 key1:value1 key2:value2 key3:value3 key (type) value (tokens)

Create a freqdist tally with list comprehension syntax 1. >>> from nltk.probability import FreqDist 2. >>> wubFD = FreqDist(word for word in text) 3. >>> wubFD.items()[:30] 4. [(u'all', 13), (u'semantic', 1), (u'pardon', 1), (u'switched', 1), (u'Kindred', 1), (u'splashing', 1), (u'excellent', 1), (u'month', 1), (u'four', 1), (u'sunk', 1), (u'straws', 1), (u'sleep', 1), (u'skin', 1), (u'go', 8), (u'meditation', 2), (u'shrugged', 1), (u'milk', 1), (u'issues', 1), (u'...."', 1), (u'apartment', 1), (u'to', 57), (u'tail', 3), (u'dejectedly', 1), (u'squeezing', 1), (u'Not', 1), (u'sorry', 2), (u'Now', 2), (u'Eat', 1), (u'fists', 1), (u'And', 5)] 24-Oct-2014NLP, Prof. Howard, Tulane University 8

A new terminology from statistics 1. >>> wubFD >>> len(wubFD.keys()) >>> len(wubFD.samples()) >>> wubFD.B() >>> sum(wubFD.values()) >>> wubFD.N() Oct-2014NLP, Prof. Howard, Tulane University 9

A dicionary maps a key to a value, an experiment observes an outcome from a sample 24-Oct-2014NLP, Prof. Howard, Tulane University 10 key1:value1 key2:value2 key3:value3 key (type) sample B value (tokens) outcome N

Value to key 1. >>> wubFD.max() 2. u'.' 3. >>> wubFD[u'.'] >>> wubFD.Nr(289) >>> wubFD.r_Nr(289) 8. defaultdict(, {0: 0, 1: 592, 2: 140, 3: 51, 4: 32, 5: 17, 6: 13, 7: 13, 8: 8, 9: 6, 10: 5, 11: 3, 12: 1, 13: 3, 14: 3, 15: 3, 17: 1, 18: 4, 19: 2, 20: 2, 21: 1, 22: 1, 23: 2, 25: 1, 26: 1, 28: 1, 30: 1, 33: 2, 34: 4, 164: 1, 37: 1, 39: 1, 41: 1, 48: 1, 53: 1, 54: 1, 56: 1, 57: 1, 59: 1, 61: 1, 66: 1, 69: 1, 289: 1, 141: 1, 146: 1}) 24-Oct-2014NLP, Prof. Howard, Tulane University 11

More new methods 1. >>> wubFD['the'] > wubFD['wub'] 2. >>> wubFD.freq('.') Oct-2014NLP, Prof. Howard, Tulane University 12

An example 1. >>> wubFD['wub'] >>> wubFD.N() from __future__ import division 6. 54/ >>> wubFD.freq('wub') >>> round(wubFD.freq('wub'), 3) Oct-2014NLP, Prof. Howard, Tulane University 13

A hapax is a word that only appears once in a text 1. >>> wubFD.hapaxes()[:30] 2. [u'...."', u'1952', u'://', u'Am', u'An', u'Anything', u'Apparently', u'Are', u'Atomic', u'BEYOND', u'Back', u'Be', u'Beyond', u'Blundell', u'By', u'Cheer', u'DICK', u'Dick', u'Distributed', u'Do', u'Earth', u'Earthmen', u'Eat', u'Eating', u'End', u'Extensive', u'Finally', u'For', u'Good', u'Greg'] 3. >>> len(wubFD.hapaxes()) >>> wubFD.Nr(1) Oct-2014NLP, Prof. Howard, Tulane University 14

How to create a table of results with tabulate() 1. >>> wubFD.tabulate(10) 2.. " the, I ' said The to." >>> wubFD.tabulate(10, 20) 5. wub it," and of you ?" It his s Oct-2014NLP, Prof. Howard, Tulane University 15

24-Oct-2014NLP, Prof. Howard, Tulane University 16 wubFD.plot(50)

24-Oct-2014NLP, Prof. Howard, Tulane University 17 wubFD.plot(50, cumulative=True)

Zipf's law   Zipf's law states that given some corpus of natural language utterances, the frequency of any word is inversely proportional to its rank in the frequency table.  Thus the most frequent word will occur approximately twice as often as the second most frequent word, three times as often as the third most frequent word, etc.  For example, in the Brown Corpus of American English text, the word "the" is the most frequently occurring word, and by itself accounts for nearly 7% of all word occurrences (69,971 out of slightly over 1 million).  True to Zipf's Law, the second-place word "of" accounts for slightly over 3.5% of words (36,411 occurrences), followed by "and" (28,852).  Only 135 vocabulary items are needed to account for half the Brown Corpus. 24-Oct-2014NLP, Prof. Howard, Tulane University 18

Other examples  The same relationship occurs in many other rankings unrelated to language, such as the population ranks of cities in various countries, corporation sizes, income rankings, ranks of number of people watching the same TV channel, and so on.  The appearance of the distribution in rankings of cities by population was first noticed by Felix Auerbach in Oct-2014NLP, Prof. Howard, Tulane University 19

What to do about the most frequent words  >>> import nltk  >>> stop=nltk.corpus.stopwords.words('english')  >>> stop  ['i', 'me', 'my', 'myself', 'we', 'our', 'ours', 'ourselves', 'you', 'your', 'yours', 'yourself', 'yourselves', 'he', 'him', 'his', 'himself', 'she', 'her', 'hers', 'herself', 'it', 'its', 'itself', 'they', 'them', 'their', 'theirs', 'themselves', 'what', 'which', 'who', 'whom', 'this', 'that', 'these', 'those', 'am', 'is', 'are', 'was', 'were', 'be', 'been', 'being', 'have', 'has', 'had', 'having', 'do', 'does', 'did', 'doing', 'a', 'an', 'the', 'and', 'but', 'if', 'or', 'because', 'as', 'until', 'while', 'of', 'at', 'by', 'for', 'with', 'about', 'against', 'between', 'into', 'through', 'during', 'before', 'after', 'above', 'below', 'to', 'from', 'up', 'down', 'in', 'out', 'on', 'off', 'over', 'under', 'again', 'further', 'then', 'once', 'here', 'there', 'when', 'where', 'why', 'how', 'all', 'any', 'both', 'each', 'few', 'more', 'most', 'other', 'some', 'such', 'no', 'nor', 'not', 'only', 'own', 'same', 'so', 'than', 'too', 'very', 's', 't', 'can', 'will', 'just', 'don', 'should', 'now'] 24-Oct-2014NLP, Prof. Howard, Tulane University 20

Q7 Conditional frequency Next time 24-Oct-2014NLP, Prof. Howard, Tulane University 21