Machine Learning biotml

From Anote2Wiki
Jump to: navigation, search

About BioTML Plug-in

BioTML is a Machine learning based plug-in that enables users to create models, trained based on available annotations (NER or RE) created by the user using the manual curation environment. These models can be used in a prediction stage to annotate documents from a selected @Note corpus.

Install Plug-in

The BioTML can be installed as any other plugin following the instructions available in:

How to Install

User HowTOs

NER Model Creation

RE Model Creation

NER Annotation of documents using machine learning models

RE Annotation of documents using machine learning models

MVC AIBench Model:

Operations:

CreateNERModelFile: Creates a BioTML NER model file using an existent NERProcess for novel entity identification in other corpus.

CreateREModelFile: Creates a BioTML RE model file using an existent REProcess for novel relations identification in other corpus.

AnnotateCorpusWithNERModel: Apply the NER Process to Corpus based in the provided machine learning model prediction for entity identification.

AnnotateCorpusWithREModel: Apply the RE Process to Corpus based in the provided machine learning model prediction associated with an existent NERProcess for entity identification.