What is it about?
This paper presents an approach and an open source tool named OntoPAWLS to allow domain experts to annotate PDF documents containing industrial procedures. The use of a procedure ontology helps to process those annotations to automatically extract semantically structured data and build procedural knowledge graph.
Featured Image
Photo by ThisisEngineering RAEng on Unsplash
Why is it important?
A lot of relevant knowledge for industrial operators is "buried" in textual documents. Artificial Intelligence promises to improve and facilitate the access to this untapped procedural knowledge, but machine learning algorithms often need large amounts of labeled training data which must be collected from domain experts. There is the need to ease the document annotation task, so to get high quality data out of textual documents, so to exploit such knowledge to support people working in industry in their daily operations.
Perspectives
I think that the field of procedural knowledge, as "how to" information, is of paramount importance in knowledge representation research, as it is more difficult to codify, formalize and digitize. This is the right moment to exploit hybrid human-machine approaches, combining people expertise and artificial intelligence, to bring innovation to the extraction and reuse of information that otherwise would remain untapped in documents.
Irene Celino
Cefriel
Read the Original
This page is a summary of: Annotation and Extraction of Industrial Procedural Knowledge from Textual Documents, December 2023, ACM (Association for Computing Machinery),
DOI: 10.1145/3587259.3627570.
You can read the full text:
Contributors
The following have contributed to this page
