nltk.parse.nonprojectivedependencyparser.DependencyScorerI

class documentation

class DependencyScorerI(object): (source)

Known subclasses: nltk.parse.nonprojectivedependencyparser.DemoScorer, nltk.parse.nonprojectivedependencyparser.NaiveBayesDependencyScorer

View In Hierarchy

A scorer for calculated the weights on the edges of a weighted dependency graph. This is used by a ProbabilisticNonprojectiveParser to initialize the edge weights of a DependencyGraph. While typically this would be done by training a binary classifier, any class that can return a multidimensional list representation of the edge weights can implement this interface. As such, it has no necessary fields.

Method __init__ Undocumented
Method score scored. :rtype: A three-dimensional list of numbers. :return: The score is returned in a multidimensional(3) list, such that the outer-dimension refers to the head, and the inner-dimension refers to the dependencies...
Method train Typically the edges present in the graphs can be used as positive training examples, and the edges not present as negative examples.
def __init__(self): (source)
overridden in nltk.parse.nonprojectivedependencyparser.NaiveBayesDependencyScorer

Undocumented

def score(self, graph): (source)
overridden in nltk.parse.nonprojectivedependencyparser.DemoScorer, nltk.parse.nonprojectivedependencyparser.NaiveBayesDependencyScorer

scored. :rtype: A three-dimensional list of numbers. :return: The score is returned in a multidimensional(3) list, such that the outer-dimension refers to the head, and the inner-dimension refers to the dependencies. For instance, scores[0][1] would reference the list of scores corresponding to arcs from node 0 to node 1. The node's 'address' field can be used to determine its number identification.

For further illustration, a score list corresponding to Fig.2 of Keith Hall's 'K-best Spanning Tree Parsing' paper:

scores = [[[], [5], [1], [1]],
[[], [], [11], [4]], [[], [10], [], [5]], [[], [8], [8], []]]

When used in conjunction with a MaxEntClassifier, each score would correspond to the confidence of a particular edge being classified with the positive training examples.

Parameters
graph:DependencyGraphA dependency graph whose set of edges need to be
def train(self, graphs): (source)
overridden in nltk.parse.nonprojectivedependencyparser.DemoScorer, nltk.parse.nonprojectivedependencyparser.NaiveBayesDependencyScorer

Typically the edges present in the graphs can be used as positive training examples, and the edges not present as negative examples.

Parameters
graphs:list(DependencyGraph)A list of dependency graphs to train the scorer.