Method and system for mapping text phrases to a taxonomy
AssigneeKnowtions Research Inc.
InventorsAlexandre Tomberg, Rohollah Soltani Bidgoli
Life of this patent
Estimating the term…
Abstract
There is provided systems and methods for mapping of text phrases to a taxonomy. In an aspect, the method including: receiving the corpus and the taxonomy; mapping the text phrases in the corpus to a set of word embeddings in a word embedding space, where each sequence of word embeddings corresponds to individual words in one of the text phrases; vectorizing the taxonomy to a set of node embeddings in a node embedding vector space; mapping the set of word embeddings to the set of node embeddings using a mapping function, the mapping function outputting points in the node embedding space associated with sequences in the word embeddings; and outputting the mapping function.
Claims12
A computer-implemented method for mapping of text phrases in a corpus of biomedical data to a biomedical taxonomy, the method comprising:
- receiving the corpus of biomedical data and the biomedical taxonomy;
- mapping the text phrases in the corpus to a set of word embeddings in a word embedding space, where each sequence of word embeddings corresponds to individual words in one of the text phrases;
- vectorizing the taxonomy to a set of node embeddings in a node embedding vector space, the taxonomy comprises a graph with concepts at each vertex and relationships between respective concepts at the edges connecting respective vertices mapping the set of word embeddings in the word embedding space to the vectorized set of node embeddings in the node embedding vector space using a mapping function between the spaces, the mapping function outputting points in the node embedding space associated with sequences in the word embeddings, the mapping function determined using a trained machine learning model, the machine learning model taking as input the set of word embeddings and the set of node embeddings and trained using phrase-concept pairs previous labelled for at least a portion of the taxonomy;
- and
- annotating text phrases in the corpus of biomedical data with the biomedical taxonomy using the mapping function, and outputting the annotated text phrases to an interface.
A system for mapping of text phrases in a corpus of biomedical data to a biomedical taxonomy, the system comprising one or more processors and memory, the memory storing the corpus and taxonomy, the one or more processors in communication with the memory and configured to execute:
- an input module to receive the corpus of biomedical data and the biomedical taxonomy;
- a corpus module to map the text phrases in the corpus to a set of word embeddings in a word embedding space, where each sequence of word embeddings corresponds to individual words in one of the text phrases;
- a taxonomy module to vectorize the taxonomy to a set of node embeddings in a node embedding vector space, the taxonomy comprises a graph with concepts at each vertex and relationships between respective concepts at the edges connecting respective vertices;
- a mapping module to map the set of word embeddings in the word embedding space to the vectorized set of node embeddings in the node embedding vector space using a mapping function between the spaces, the mapping function outputting points in the node embedding space associated with sequences in the word embeddings, the mapping function determined using a trained machine learning model, the machine learning model taking as input the set of word embeddings and the set of node embeddings and trained using phrase-concept pairs previous labelled for at least a portion of the taxonomy, and to annotate text phrases in the corpus of biomedical data with the biomedical taxonomy using the mapping function;
- and
- an output module to output the annotated text phrases to an interface.
Dependencies are inferred from the claim text.