This is my homework 1 from CS6320 in the University of Texas at Dallas, Spring 2018
- environment: Python 3
- package used: nltk, pandas
- put all files in the same folder: homework1.py, corpus.txt(or any .txt as the word training set)
python homework1.py 'corpus.txt' "Apple computer is the first product of the company" "Apple introduced the new version of iPhone in 2008"
- case sensitive, ex:”iPhone” != “iphone”. There in Sentence 2 was an “iphone”, I change it to “iPhone” for a better result.
- The set size to be compared is sentence length minus one, for the reason to simply the calculation, so I ignore the start sign(denoted as ‘^’) and end sign(denoted as ‘$’)
- When bigram probability without smoothing, if the word of sentence never show up in the training set, to prevent dividing by zero, I skip this bigram and list how many bigrams are skipped by this case.
JCMacBook:CS6320_NLP JC$ python homework1.py 'corpus.txt' "Apple computer is the first product of the company" "Apple introduced the new version of iPhone in 2008"
bigram counts table for Sentence 1
Apple computer is the first product of company
Apple 0 0 7 1 0 0 0 0
computer 0 0 0 0 0 0 0 0
is 0 0 0 5 0 0 0 0
the 23 0 0 0 28 1 0 43
first 2 1 0 0 0 1 1 0
product 0 0 0 0 0 0 0 0
of 17 2 0 85 0 0 0 2
company 0 0 0 0 0 0 0 0
bigram counts table for Sentence 2
Apple introduced the new version of iPhone in 2008
Apple 0 20 1 0 0 0 0 6 0
introduced 1 0 15 2 0 0 0 0 0
the 23 0 0 6 0 0 29 0 0
new 1 0 0 0 0 0 1 0 0
version 0 0 0 0 0 3 0 0 0
of 17 0 85 0 0 0 1 0 0
iPhone 0 0 0 0 0 0 0 0 0
in 8 0 55 0 0 0 1 0 1
2008 0 0 0 0 0 0 0 0 0
bigram probability table for Sentence 1
Apple computer is the first product of company
Apple 0.000000 0.000000 0.12963 0.001326 0.000000 0.000000 0.000000 0.000000
computer 0.000000 0.000000 0.00000 0.000000 0.000000 0.000000 0.000000 0.000000
is 0.000000 0.000000 0.00000 0.006631 0.000000 0.000000 0.000000 0.000000
the 0.051111 0.000000 0.00000 0.000000 0.571429 0.032258 0.000000 0.551282
first 0.004444 0.058824 0.00000 0.000000 0.000000 0.032258 0.002674 0.000000
product 0.000000 0.000000 0.00000 0.000000 0.000000 0.000000 0.000000 0.000000
of 0.037778 0.117647 0.00000 0.112732 0.000000 0.000000 0.000000 0.025641
company 0.000000 0.000000 0.00000 0.000000 0.000000 0.000000 0.000000 0.000000
bigram probability table for Sentence 2
Apple introduced the new version of iPhone in 2008
Apple 0.000000 0.526316 0.001326 0.000000 0.0 0.000000 0.000000 0.020548 0.000000
introduced 0.002222 0.000000 0.019894 0.041667 0.0 0.000000 0.000000 0.000000 0.000000
the 0.051111 0.000000 0.000000 0.125000 0.0 0.000000 0.475410 0.000000 0.000000
new 0.002222 0.000000 0.000000 0.000000 0.0 0.000000 0.016393 0.000000 0.000000
version 0.000000 0.000000 0.000000 0.000000 0.0 0.008021 0.000000 0.000000 0.000000
of 0.037778 0.000000 0.112732 0.000000 0.0 0.000000 0.016393 0.000000 0.000000
iPhone 0.000000 0.000000 0.000000 0.000000 0.0 0.000000 0.000000 0.000000 0.000000
in 0.017778 0.000000 0.072944 0.000000 0.0 0.000000 0.016393 0.000000 0.090909
2008 0.000000 0.000000 0.000000 0.000000 0.0 0.000000 0.000000 0.000000 0.000000
bigram counts table for Sentence 1 with smoothing
Apple computer is the first product of company
Apple 1 1 8 2 1 1 1 1
computer 1 1 1 1 1 1 1 1
is 1 1 1 6 1 1 1 1
the 24 1 1 1 29 2 1 44
first 3 2 1 1 1 2 2 1
product 1 1 1 1 1 1 1 1
of 18 3 1 86 1 1 1 3
company 1 1 1 1 1 1 1 1
bigram counts table for Sentence 2 with smoothing
Apple introduced the new version of iPhone in 2008
Apple 1 21 2 1 1 1 1 7 1
introduced 2 1 16 3 1 1 1 1 1
the 24 1 1 7 1 1 30 1 1
new 2 1 1 1 1 1 2 1 1
version 1 1 1 1 1 4 1 1 1
of 18 1 86 1 1 1 2 1 1
iPhone 1 1 1 1 1 1 1 1 1
in 9 1 56 1 1 1 2 1 2
2008 1 1 1 1 1 1 1 1 1
bigram probability table for Sentence 1 with smoothing
Apple computer is the first product of company
Apple 0.000259 0.000292 0.002307 0.000480 0.000289 0.000290 0.000264 0.000286
computer 0.000259 0.000292 0.000288 0.000240 0.000289 0.000290 0.000264 0.000286
is 0.000259 0.000292 0.000288 0.001440 0.000289 0.000290 0.000264 0.000286
the 0.006213 0.000292 0.000288 0.000240 0.008377 0.000581 0.000264 0.012604
first 0.000777 0.000583 0.000288 0.000240 0.000289 0.000581 0.000528 0.000286
product 0.000259 0.000292 0.000288 0.000240 0.000289 0.000290 0.000264 0.000286
of 0.004660 0.000875 0.000288 0.020638 0.000289 0.000290 0.000264 0.000859
company 0.000259 0.000292 0.000288 0.000240 0.000289 0.000290 0.000264 0.000286
bigram probability table for Sentence 2 with smoothing
Apple introduced the new version of iPhone in 2008
Apple 0.000259 0.006085 0.000480 0.000289 0.000292 0.000264 0.000288 0.001889 0.000292
introduced 0.000518 0.000290 0.003840 0.000867 0.000292 0.000264 0.000288 0.000270 0.000292
the 0.006213 0.000290 0.000240 0.002023 0.000292 0.000264 0.008636 0.000270 0.000292
new 0.000518 0.000290 0.000240 0.000289 0.000292 0.000264 0.000576 0.000270 0.000292
version 0.000259 0.000290 0.000240 0.000289 0.000292 0.001056 0.000288 0.000270 0.000292
of 0.004660 0.000290 0.020638 0.000289 0.000292 0.000264 0.000576 0.000270 0.000292
iPhone 0.000259 0.000290 0.000240 0.000289 0.000292 0.000264 0.000288 0.000270 0.000292
in 0.002330 0.000290 0.013439 0.000289 0.000292 0.000264 0.000576 0.000270 0.000584
2008 0.000259 0.000290 0.000240 0.000289 0.000292 0.000264 0.000288 0.000270 0.000292
the total probabilities for Sentence 1
did multiplied: 8 compare length: 8
ignored 0 probabilities due to zero frequency
0.0
the total probabilities for Sentence 2
did multiplied: 8 compare length: 8
ignored 0 probabilities due to zero frequency
0.0
the total probabilities for Sentence 1 with smoothing
did multiplied: 8 compare length: 8
4.045759371642444e-23
the total probabilities for Sentence 2 with smoothing
did multiplied: 8 compare length: 8
1.323528368268833e-24
Sentence 1 is better, which is " Apple computer is the first product of the company "
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent