bugfinder.features.extraction.node2vec.model
- class bugfinder.features.extraction.node2vec.model.Node2VecModel(dataset, deprecation_warning=None)
Bases:
AbstractProcessingClass for process the dataset and train a word2vec model using the node2vec algorithm to generate the corpus used as input
- algorithm = 1
- execute(name, **kwargs)
Run the processing. This function receives the processed dataset, retrieves all edges related to data and control flow (REACHES and FLOWS_TO), generates the graphs, run the node2vec algorithm to generate the random walks in the graphs, creates the model and saves it.
- Parameters
name (str) – This parameter will be the name of the model saved in disk.
- min_count = 1
- num_walks = 10
- p = 1
- q = 1
- seed = None
- tokens = {}
- vector_length = 128
- walk_length = 50
- window_dim = 10
- workers = 4