Commit 42d66c63 authored by Victor Zimmermann's avatar Victor Zimmermann
Browse files

Merge branch 'colour_dist' into 'master' 

Changed binary colouring to colour distribution and better Statistics

See merge request zimmermann/absinth!1
parents 86dc9855 72463d07

src/absinth.py

+26 −12
Changes for src/absinth.py: 26 added lines, 12 removed lines.
Original line number Diff line number Diff line
@@ -564,15 +564,25 @@ def colour_graph(graph: nx.Graph, root_hub_list: list) -> nx.Graph: 
    

    

    for node in graph.nodes:

        

        graph.node[node]['dist'] = [0] * len(root_hub_list)

            

        if node in root_hub_list:

            graph.node[node]['sense'] = root_hub_list.index(node)

            

            root_idx = root_hub_list.index(node)

            

            graph.node[node]['sense'] = root_idx

            graph.node[node]['dist'][root_idx] = 1

            

        else:

            

            graph.node[node]['sense'] = None

    

    max_iteration_count = config.max_colour_iteration_count

    

    iteration_count = 0

    stable = False

    

    while stable == False and iteration_count <= max_iteration_count:

        

        graph_copy = deepcopy(graph)
@@ -591,10 +601,14 @@ def colour_graph(graph: nx.Graph, root_hub_list: list) -> nx.Graph: 
                    neighbor_weight_list[graph_copy.node[neighbor]['sense']] \

                        += 1 - graph_copy[node][neighbor]['weight']

            

            

            if any(neighbor_weight_list):

                

                graph.node[node]['dist'] = np.mean([graph.node[node]['dist'],

                                                    neighbor_weight_list], axis=0)

                

                old_colour = graph_copy.node[node]['sense']

                new_colour = np.argmax(neighbor_weight_list)

                new_colour = np.argmax(graph.node[node]['dist'])

                

                if old_colour != new_colour:

                    stable = False
@@ -651,22 +665,23 @@ def disambiguate_colour(graph: nx.Graph, root_hub_list: list, context_list: list 
            

            if text in coloured_graph.nodes:

                

                text_colour = coloured_graph.node[text]['sense']

                text_colour_dist = coloured_graph.node[text]['dist']

                

                if text_colour == None:

                if not any(text_colour_dist):

                    

                    pass

                

                else:

                    

                    text_root = root_hub_list[text_colour]

                    for root_hub in root_hub_list:

                        

                    if nx.has_path(coloured_graph , text, text_root):

                        root_hub_idx = root_hub_list.index(root_hub)

                    

                        if nx.has_path(coloured_graph , text, root_hub):

                            

                            shortest_path = nx.shortest_path(coloured_graph ,

                                                            text,

                                                        root_hub_list[text_colour],

                                                            root_hub,

                                                            'weight')

                            total_weight = 0
@@ -676,8 +691,8 @@ def disambiguate_colour(graph: nx.Graph, root_hub_list: list, context_list: list 
                                total_weight += \

                                    coloured_graph[sub_from][sub_to]['weight']

                

            

                        score[text_colour] += 1/(1+total_weight)

                            score[root_hub_idx] += (1/(1+total_weight)) \

                             * colour_graph.node[text]['dist'][root_hub_idx]

                    

                        else:

                            pass
@@ -768,7 +783,7 @@ def disambiguate_mst(graph: nx.Graph, root_hub_list: list, 
                pass

        

        # If disambiguator does not detect a sense, return singleton.

        if any(score_array):

        if not any(score_array):

            

            pass
@@ -801,8 +816,7 @@ def main(topic_id: int, topic_name: str, result_dict: dict) -> None: 
    """

    

    print('[a]', 'Inducing word senses for {}.'.format(topic_name))

    graph, root_hub_list, stat_dict = induce(topic_name,

                                                              result_dict[topic_id])

    graph, root_hub_list, stat_dict = induce(topic_name,result_dict[topic_id])

    

    colour_rank = config.colour_rank

    mst_rank = config.mst_rank

src/config.py

+12 −1
Changes for src/config.py: 12 added lines, 1 removed line.
Original line number Diff line number Diff line
@@ -11,6 +11,18 @@ dataset = "../WSI-Evaluator/datasets/dataset/" 
test = "../WSI-Evaluator/datasets/trial/"

output = "../output/"



'''

Disambiguation Pipeline

There are multiple disambiguation methods implemented. Specify the order in

which they should be merged and whether or not conflicts should be resolved.

If conflicts are not resolved, the first method with a positive result is used.

Methods labeled with 0 are ignored.

At least one method must be given a value != 0.

'''

resolve_conflicts = False #not yet implemented

mst_rank = 0

colour_rank = 1



'''

Choose stop words and allowed pos-tags.

- Stop words will not be considered for nodes.
@@ -54,5 +66,4 @@ lemma = False 
'''

colouring options

'''

use_colouring = True

max_colour_iteration_count = 50