If not locked, an unknown name is added to the vocabulary. If locked, return -1 on OOV. @param featname @return

Does feature extraction on one sentence. Input: textual representation of sentence Output: fills up modelSentence with numberized features

Please only use absolute paths, e.g. /cmu/arktweetnlp/6mpaths e.g. http://stackoverflow.com/questions/1464291/how-to-really-read-text-file-from-class

Get the text from a raw Tweet JSON string. @param tweetJson @return null if there is no text field, or invalid JSON.

This is intended for raw tweet text -- we do some HTML entity unescaping before running the tagger. This function normalizes the input text BEFORE ca

Adds into labelScores

Computes unnormalized log-potentials. CLOBBERS labelScores

Adds into labelScores

Runs the correct algorithm (make config option perhaps)

THIS CLOBBERS THE LABELS, stores its decoding into them. Does progressive rolling edge feature extraction

"given labels" i.e. at trainingtime labels are observed. You hide the current one and predict it given you know the previous. So you get funny increme

Loads a model from a file. The tagger should be ready to tag after calling this. @param modelFilename @throws IOException

Try to get a file, if it doesn't exist, backoff to a resource. @throws IOException

assume mSent's labels hold the tagging.

Implementation of the (Michelot 1986) technique for projecting a real vector into the feasible region K defined by the following constraints: - comp

Input: sentence Output: labelIndexes, featureIDs/Values through positionFeaturePairs We want to yield a sequence of (t, featID, featValue) pairs, to