summaryrefslogtreecommitdiff
path: root/spellchecker
AgeCommit message (Collapse)AuthorFilesLines
2015-04-02Consider the probability of the whole sentencePeter Wu1-15/+40
2015-04-02SpellCorrector: look for differences in likelihoodPeter Wu1-0/+4
2015-04-02Calculate probability for unigramPeter Wu1-1/+1
Find L * P(word) instead of L * #word.
2015-04-02Fix probability calculationsPeter Wu2-36/+45
Replace monkey patching by an actual implementation that correctly combines channel probability (language model), prior and interpolation.
2015-04-02Support continuously processing stdinPeter Wu2-1/+20
2015-04-01Use Kerninghans confusion matrixPeter Wu1-1573/+1261
2015-04-01SpellChecker: FUBAREDPeter Wu1-31/+34
This is the submitted version, it contains random attempts and a fix (actually not that sure that this is the submitted version...). The Double.MIN_VALUE fix is probably missing.
2015-04-01Netbeans: enable assertions on runPeter Wu1-1/+1
2015-04-01Rate multi n-grams higherPeter Wu1-4/+16
2015-04-01Set true for peachPeter Wu1-1/+1
2015-04-01Count unigramsPeter Wu1-0/+9
2015-04-01Initial attempt at rating sentencesPeter Wu1-6/+145
2015-04-01Do not include unchanged wordsPeter Wu2-1/+6
2015-04-01Remove unused calculateChannelModelProbabilityPeter Wu1-8/+0
Channel probability is already calculated during getCandidateWords.
2015-04-01Calculate noisy channel probabilityPeter Wu2-0/+29
2015-04-01Replace Set by Map for later extensionPeter Wu2-8/+13
2015-04-01Immediately return new listPeter Wu1-1/+1
The filtering is already done.
2015-03-31Refactor getCandidateWords into separate methodsPeter Wu1-5/+23
For readability and testability.
2015-03-31Refactor to improve understandabilityPeter Wu1-19/+25
Split deletion and substitution, introduce consistent style (comment, example, implementation).
2015-03-31Refactor getCandidateWords for ConfusionMatrixPeter Wu3-18/+70
2015-03-31Implement Damerau-Levenshtein distance (add transposition)Peter Wu2-2/+23
Suggested around 6:16 of the Coursera lecture on "The Noisy Channel Model of Spelling". https://class.coursera.org/nlp/lecture/22
2015-03-31Add simple getSmoothedCount implementationPeter Wu1-3/+7
2015-03-31Tests and implementation for getCandidateWordsPeter Wu2-5/+155
SpellCorrector.ALPHABET is made static to allow for the test.
2015-03-30Reformat code (apply consistent coding style)Peter Wu4-132/+113
2015-03-30Initial checkin of assignment4-0.0Peter Wu15-0/+566662
Extracted from assignment4-0.0.zip (without build/ dirs). The CRLF in test-sentences.txt was replaced by LFs. https://www.win.tue.nl/~wstahw/edu/2ID90/assignments/4/assignment4-0.0.zip size 5189037 sha1 5e0a58acffbcac6e5f13837c98c2c0bb936304b5