v0.1.0

Added

• Modules mowl.ontology.extend and mowl.ontology.create created including insert_annotations and create_from_triples methods, respectively.
• Package deprecated as dependency.
• ELEmbeddingModel abstract class that contains basic functionality for models embedding the EL language.
• Implementation of ELBoxEmbeddings
• ELDataset class to work with EL models.
• Module mowl.nn where torch.nn-based modules will reside. The first module existing there is the abstract module for EL models.
• Module mowl.models where implementation of ELEmbeddings and ELBoxEmbeddings reside.
• PyKEEN as dependency
• GDAHumanELDataset and GDAMouseELDataset, which are reduced versions of GDAHumanDataset and GDAMouseDataset, respectively. The new datasets can be normalized into the EL language using the jcel library.
• Started implementation of unit tests for datasets and walking modules.

Changed

• All builtin datasets can be imported from mowl.datasets.builtin module.
• Updated implementation of ELEmbeddings
• Changed method mowl.datasets.PathDataset.get_evaluation_classes to property mowl.datasets.PathDataset.evaluation_classes
• #25 Walking methods now accept an optional parameter for a list of nodes to filter the random walks.

Deprecated

• Modules mowl.datasets.ppi_yeast and mowl.datasets.gda.
• File mowl.datasets.build_ontology.
• Class mowl.embeddings.elembeddings.ELEmbeddings. Future versions will point to mowl.models.ELEmbeddings
• Method mowl.corpus.base.extract_annotation_corpus. Future versions will split this method into two: mowl.corpus.base.extract_annotation_corpus and mowl.corpus.base.extract_and_save_annotation_corpus.
• Method mowl.datasets.PathDataset.get_evaluation_classes.