This is a multi-ontology concept recognition model. It takes OBO ontologies as inputs, and trains a text classifier for identifying specific ontology terms in text.
-
python==3.9.12
-
tensorflow==2.9.1
-
tensorflow-addons==0.18.0
-
regex
-
numpy
-
A copy of the pre-trained elmo model
Example training parameter json:
{
"model_type": "sae",
"obo_params" : "path/to/obo/parameter/file.json",
"hub_dir": "directory/containing/elmo/model",
"output": "directory/for/trained/model",
"num_filters": 1024,
"concept_dim": 1024,
"compression_ratio": 8,
"epochs": 10000,
"mean": true
}
Example OBO parameter json:
{
"ontologies": [
{ "name": "Human Phenotype Ontology",
"obo_path": "ontologies/hpo.obo",
"root_id": "HP:0000001",
"main": 1
},
{ "name": "Mammalian Phenotype Ontology",
"obo_path": "ontologies/mp.obo",
"root_id": "MP:0000001",
"main": 0
}
]
}
To train model:
python adorNER/train.py --params path/to/training/parameter/file.json
Example annotation parameter json:
{
"hub_dir": "elmo",
"model_dir": "directory/of/trained/model",
"input_dir": "directory/containing/txt/files",
"output_dir": "directory/to/output/identified/ontology/terms",
"threshold": 0.55,
"max_rows": -1
}
To use model:
python adorNER/annotate_text.py --params path/to/annotation/parameter/file.json
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent