wiki:en/AdvancedNlpCourse2015/AutomaticCorrection

Automatic language correction

IA161 Advanced NLP Course, Course Guarantee: Aleš Horák

Prepared by: Ján Švec

State of the Art

Language correction nowadays has many potential applications on large amount of informal and unedited text generated online, among other things: web forums, tweets, blogs, and email. Automatic language correction can consist of many areas including: spell checking, grammar checking and word completion.

In the theoretical lesson we will introduce and compare various methods to automatically propose and choose a correction for an incorrectly written word. Spell checking is the process of detecting and sometimes providing spelling suggestions for incorrectly spelled words in a text. The lesson will also focus on grammatical checking problems, which are the most difficult and complex type of language errors, because grammar is made up of a very extensive number of rules and exceptions. We will also say a few words about word completion.

The lesson will also answer a question "How difficult is to develop a spell-checker?". And also describe a system that performs spell-checking and autocorrection.

References

  1. CHOUDHURY, Monojit, et al. "How Difficult is it to Develop a Perfect Spell-checker? A Cross-linguistic Analysis through Complex Network Approach" Graph-Based Algorithms for Natural Language Processing, pages 81–88, Rochester, 2007. Source
  2. WHITELAW, Casey, et al. "Using the Web for Language Independent Spellchecking and Autocorrection" Proceedings of the 2009 Conference on Empirical Methods in Natural Language Processing, pages 890–899, Singapore, 2009. Source
  3. GUPTA, Neha, MATHUR, Pratistha. "Spell Checking Techniques in NLP: A Survey" International Journal of Advanced Research in Computer Science and Software Engineering, volume 2, issue 12, pages 217-221, 2012. Source
  4. HLADEK, Daniel, STAS, Jan, JUHAR, Jozef. "Unsupervised Spelling Correction for the Slovak Text." Advances in Electrical and Electronic Engineering 11 (5), pages 392-397, 2013. Source

Slides

http://nlp.fi.muni.cz/trac/research/raw-attachment/wiki/en/AdvancedNlpCourse/anlp-14-AutomaticCorrection.pdf

Practical Session

In theoretical lesson we have become acquainted with various approaches how spelling correctors work. Now we will get to know how a simple spellchecker based on edit distance works.

The example is based on Peter Norvig's Spelling Corrector in python. The spelling corrector will be trained with a large text file consisting of about a million words.

We will test this tool on prepared data. Your goal will be to enhance spellchecker's accuracy. If you finish early, there is a bonus question in the task section.

  1. Download prepared script spell.py and training data collection big.txt.
  2. Test the script python ./spell.py in your working directory.
  3. Open it in your favourite editor and we will walk through its functionality.

Spellchecker functionality with examples

  1. Spellchecker is trained from file big.txt which is a concatenation of several public domain books from Project Gutenberg and lists of most frequent words from Wiktionary and the British National Corpus. Function train stores how many times each word occurs in the text file. NWORDS[w] holds a count of how many times the word w has been seen.
def words(text): return re.findall('[a-z]+', text.lower()) 

def train(features):
    model = collections.defaultdict(lambda: 1)
    for f in features:
        model[f] += 1
    return model

NWORDS = train(words(file('big.txt').read()))
  1. Edit distance 1 is represented as function edits1 - it represents deletion (remove one letter), a transposition (swap adjacent letters), an alteration (change one letter to another) or an insertion (add a letter). For a word of length n, there will be n deletions, n-1 transpositions, 26n alterations, and 26(n+1) insertions, for a total of 54n+25. Example: len(edits1('something')) = 494 words.
def edits1(word):
   splits     = [(word[:i], word[i:]) for i in range(len(word) + 1)]
   deletes    = [a + b[1:] for a, b in splits if b]
   transposes = [a + b[1] + b[0] + b[2:] for a, b in splits if len(b)>1]
   replaces   = [a + c + b[1:] for a, b in splits for c in alphabet if b]
   inserts    = [a + c + b     for a, b in splits for c in alphabet]
   return set(deletes + transposes + replaces + inserts)
  1. Edit distance 2(edits2) - applied edits1 to all the results of edits1. Example: len(edits2('something')) = 114 324 words, which is a high number. To enhance speed we can only keep the candidates that are actually known words (known_edits2). Now known_edits2('something') is a set of just 4 words: {'smoothing', 'seething', 'something', 'soothing'}.
  1. The function correct chooses as the set of candidate words the set with the shortest edit distance to the original word.
    def known(words): return set(w for w in words if w in NWORDS)
    
    def correct(word):
        candidates = known([word]) or known(edits1(word)) or known_edits2(word) or [word]
        return max(candidates, key=NWORDS.get)
    
  1. For evaluation there are prepared two test sets - developement(test1) and final test set(test2).

Task

  1. Create <YOUR_FILE>, a text file named ia161-UCO-14.txt where UCO is your university ID.
  1. Run spell.py with developement and final test sets (test1 and test2), write the results in <YOUR_FILE>.
  1. Explain the given results in few words and write it in <YOUR_FILE>.
  1. Modify the code of spell.py to increase accuraccy by 10 %. Write your new accuracy results to <YOUR_FILE>.
  1. Run the script with verbose=True and examine given results. Try to suggest at least one adjustment how to enhance spellchecker's accuracy. Write your suggestions to <YOUR_FILE>.

-Bonus question- How could you make the implementation faster without changing the results? Write your suggestions to <YOUR_FILE>.

Upload <YOUR_FILE> and edited spell.py

Do not forget to upload your resulting files to the homework vault (odevzdávárna).

Last modified 2 years ago Last modified on Sep 11, 2017 4:38:48 PM

Attachments (3)