def parse(self, tokens, tags):
"""
Parses a list of tokens in accordance to the MST parsing algorithm
for non-projective dependency parses. Assumes that the tokens to
be parsed have already been tagged and those tags are provided. Various
scoring methods can be used by implementing the C{DependencyScorerI}
interface and passing it to the training algorithm.
:type tokens: A list of C{String}.
:param tokens: A list of words or punctuation to be parsed.
:type tags: A List of C{String}.
:param tags: A list of tags corresponding by index to the words in the tokens list.
"""
self.inner_nodes = {}
# Initialize g_graph
g_graph = DependencyGraph()
for index, token in enumerate(tokens):
g_graph.nodelist.append({'word':token, 'tag':tags[index], 'deps':[], 'rel':'NTOP', 'address':index+1})
# Fully connect non-root nodes in g_graph
g_graph.connect_graph()
original_graph = DependencyGraph()
for index, token in enumerate(tokens):
original_graph.nodelist.append({'word':token, 'tag':tags[index], 'deps':[], 'rel':'NTOP', 'address':index+1})
# Initialize b_graph
b_graph = DependencyGraph()
b_graph.nodelist = []
# Initialize c_graph
c_graph = DependencyGraph()
c_graph.nodelist = [{'word':token, 'tag':tags[index], 'deps':[],
'rel':'NTOP', 'address':index+1}
for index, token in enumerate(tokens)]
# Assign initial scores to g_graph edges
self.initialize_edge_scores(g_graph)
print self.scores
# Initialize a list of unvisited vertices (by node address)
unvisited_vertices = [vertex['address'] for vertex in c_graph.nodelist]
# Iterate over unvisited vertices
nr_vertices = len(tokens)
betas = {}
while(len(unvisited_vertices) > 0):
# Mark current node as visited
current_vertex = unvisited_vertices.pop(0)
print 'current_vertex:', current_vertex
# Get corresponding node n_i to vertex v_i
current_node = g_graph.get_by_address(current_vertex)
print 'current_node:', current_node
# Get best in-edge node b for current node
best_in_edge = self.best_incoming_arc(current_vertex)
betas[current_vertex] = self.original_best_arc(current_vertex)
print 'best in arc: ', best_in_edge, ' --> ', current_vertex
# b_graph = Union(b_graph, b)
for new_vertex in [current_vertex, best_in_edge]:
b_graph.add_node({'word':'TEMP', 'deps':[], 'rel': 'NTOP', 'address': new_vertex})
b_graph.add_arc(best_in_edge, current_vertex)
# Beta(current node) = b - stored for parse recovery
# If b_graph contains a cycle, collapse it
cycle_path = b_graph.contains_cycle()
if cycle_path:
# Create a new node v_n+1 with address = len(nodes) + 1
new_node = {'word': 'NONE', 'deps':[], 'rel': 'NTOP', 'address': nr_vertices + 1}
# c_graph = Union(c_graph, v_n+1)
c_graph.add_node(new_node)
# Collapse all nodes in cycle C into v_n+1
self.update_edge_scores(new_node, cycle_path)
self.collapse_nodes(new_node, cycle_path, g_graph, b_graph, c_graph)
for cycle_index in cycle_path:
c_graph.add_arc(new_node['address'], cycle_index)
# self.replaced_by[cycle_index] = new_node['address']
self.inner_nodes[new_node['address']] = cycle_path
# Add v_n+1 to list of unvisited vertices
unvisited_vertices.insert(0, nr_vertices + 1)
# increment # of nodes counter
nr_vertices += 1
# Remove cycle nodes from b_graph; B = B - cycle c
for cycle_node_address in cycle_path:
b_graph.remove_by_address(cycle_node_address)
print 'g_graph:\n', g_graph
print
print 'b_graph:\n', b_graph
print
print 'c_graph:\n', c_graph
print
print 'Betas:\n', betas
print 'replaced nodes', self.inner_nodes
print
#Recover parse tree
print 'Final scores:\n', self.scores
print 'Recovering parse...'
for i in range(len(tokens) + 1, nr_vertices + 1):
betas[betas[i][1]] = betas[i]
print 'Betas: ', betas
new_graph = DependencyGraph()
for node in original_graph.nodelist:
node['deps'] = []
for i in range(1, len(tokens) + 1):
# print i, betas[i]
original_graph.add_arc(betas[i][0], betas[i][1])
# print original_graph
return original_graph
print 'Done.'
def parse(self, tokens, tags):
"""
Parses a list of tokens in accordance to the MST parsing algorithm
for non-projective dependency parses. Assumes that the tokens to
be parsed have already been tagged and those tags are provided. Various
scoring methods can be used by implementing the ``DependencyScorerI``
interface and passing it to the training algorithm.
:type tokens: list(str)
:param tokens: A list of words or punctuation to be parsed.
:type tags: list(str)
:param tags: A list of tags corresponding by index to the words in the tokens list.
"""
self.inner_nodes = {}
# Initialize g_graph
g_graph = DependencyGraph()
for index, token in enumerate(tokens):
g_graph.nodelist.append({
'word': token,
'tag': tags[index],
'deps': [],
'rel': 'NTOP',
'address': index + 1
})
# Fully connect non-root nodes in g_graph
g_graph.connect_graph()
original_graph = DependencyGraph()
for index, token in enumerate(tokens):
original_graph.nodelist.append({
'word': token,
'tag': tags[index],
'deps': [],
'rel': 'NTOP',
'address': index + 1
})
# Initialize b_graph
b_graph = DependencyGraph()
b_graph.nodelist = []
# Initialize c_graph
c_graph = DependencyGraph()
c_graph.nodelist = [{
'word': token,
'tag': tags[index],
'deps': [],
'rel': 'NTOP',
'address': index + 1
} for index, token in enumerate(tokens)]
# Assign initial scores to g_graph edges
self.initialize_edge_scores(g_graph)
print(self.scores)
# Initialize a list of unvisited vertices (by node address)
unvisited_vertices = [vertex['address'] for vertex in c_graph.nodelist]
# Iterate over unvisited vertices
nr_vertices = len(tokens)
betas = {}
while (len(unvisited_vertices) > 0):
# Mark current node as visited
current_vertex = unvisited_vertices.pop(0)
print('current_vertex:', current_vertex)
# Get corresponding node n_i to vertex v_i
current_node = g_graph.get_by_address(current_vertex)
print('current_node:', current_node)
# Get best in-edge node b for current node
best_in_edge = self.best_incoming_arc(current_vertex)
betas[current_vertex] = self.original_best_arc(current_vertex)
print('best in arc: ', best_in_edge, ' --> ', current_vertex)
# b_graph = Union(b_graph, b)
for new_vertex in [current_vertex, best_in_edge]:
b_graph.add_node({
'word': 'TEMP',
'deps': [],
'rel': 'NTOP',
'address': new_vertex
})
b_graph.add_arc(best_in_edge, current_vertex)
# Beta(current node) = b - stored for parse recovery
# If b_graph contains a cycle, collapse it
cycle_path = b_graph.contains_cycle()
if cycle_path:
# Create a new node v_n+1 with address = len(nodes) + 1
new_node = {
'word': 'NONE',
'deps': [],
'rel': 'NTOP',
'address': nr_vertices + 1
}
# c_graph = Union(c_graph, v_n+1)
c_graph.add_node(new_node)
# Collapse all nodes in cycle C into v_n+1
self.update_edge_scores(new_node, cycle_path)
self.collapse_nodes(new_node, cycle_path, g_graph, b_graph,
c_graph)
for cycle_index in cycle_path:
c_graph.add_arc(new_node['address'], cycle_index)
# self.replaced_by[cycle_index] = new_node['address']
self.inner_nodes[new_node['address']] = cycle_path
# Add v_n+1 to list of unvisited vertices
unvisited_vertices.insert(0, nr_vertices + 1)
# increment # of nodes counter
nr_vertices += 1
# Remove cycle nodes from b_graph; B = B - cycle c
for cycle_node_address in cycle_path:
b_graph.remove_by_address(cycle_node_address)
print('g_graph:\n', g_graph)
print()
print('b_graph:\n', b_graph)
print()
print('c_graph:\n', c_graph)
print()
print('Betas:\n', betas)
print('replaced nodes', self.inner_nodes)
print()
#Recover parse tree
print('Final scores:\n', self.scores)
print('Recovering parse...')
for i in range(len(tokens) + 1, nr_vertices + 1):
betas[betas[i][1]] = betas[i]
print('Betas: ', betas)
new_graph = DependencyGraph()
for node in original_graph.nodelist:
node['deps'] = []
for i in range(1, len(tokens) + 1):
# print i, betas[i]
original_graph.add_arc(betas[i][0], betas[i][1])
# print original_graph
return original_graph
print('Done.')
