Abner Filter

Go to Product
This Node Is Deprecated — This node is kept for backwards-compatibility, but the usage in new workflows is no longer recommended. The documentation below might contain more information.

Filters all terms contained in the given bag of words (input table) with biomedical named entities (BNER) tags assigned, not specified in the dialog. The terms are filtered out of the bag of words and the documents as well if deep preprocessing is checked.
To detect biomedical named entities, the ABNER (A Biomedical Named Entity Recognizer) Software is used (see http://pages.cs.wisc.edu/~bsettles/abner/ for details).
A term is kept if at least one of its assigned BNER tags is a valid tag. If strict filtering is set, all assigned BNER tags of a term have to be valid.


Preprocessing options

Strict filtering
If strict filtering is checked, terms need to have all specified tags applied. If not checked, at least one tag has to be applied.
Filter matching
If checked, terms which have the specified tags are filtered, otherwise they are kept. If strict filtering is checked, all tags must match to filter or keep a term.
The tags to filter for.

Deep preprocessing options

Deep preprocessing
If deep preprocessing is checked, the terms contained inside the documents are preprocessed too, this means that the documents themselves are changed too, which is more time consuming.
Document column
Specifies the column containing the documents to preprocess.
Append unchanged documents
If checked, the documents contained in the specified "Original Document column" are appended unchanged even if deep preprocessing is checked. This helps to keep the original documents in the output data table without the agonizing pain of joining.
Original Document column
Specifies the column containing the original documents which can be attached unchanged.
Ignore unmodifiable tag
If checked unmodifiable terms will be preprocessed too.

Input Ports

The input table which contains the terms to filter.

Output Ports

The output table which contains the filtered terms.

Popular Predecessors

  • No recommendations found

Popular Successors

  • No recommendations found


This node has no views




You want to see the source code for this node? Click the following button and we’ll use our super-powers to find it for you.