Interactive structure annotation with artificial intelligence
AssigneeInternational Business Machines Corporation
InventorsXin Ru Wang, Douglas Ronald Burdick, Yunyao Li
Life of this patent
Estimating the term…
Abstract
A computer system, product, and method are provided to utilize machine learning to facilitate document processing. A document collection is introduced to an artificial neural network (ANN), which subjects the document collection to table region identification within discretized contiguous areas. The documents are assigned to one or more clusters responsive to the leveraged ANN. Documents are selectively evaluated from the clusters, and one or more label corrections are applied to the ANN. The ANN generates an updated document collection incorporating the applied one or more label corrections.
Claims20
A computer system comprising:
- a processor operatively coupled to memory;
- an artificial intelligence (AI) platform, in communication with the processor, having one or more machine learning (ML) tools, the tools comprising:
- a document manager configured to cause an artificial neural network (ANN) to subject a document collection to table region identification within one or more discretized contiguous areas;
- a cluster manager configured to subject the document collection to clustering, including:
- leverage the ANN to extract one or more structures and location of the one or more structures in the document collection;
- and
- assign documents within the document collection to one or more clusters responsive to the leveraged ANN, wherein each cluster includes one or more documents having a content characteristic;
- an evaluator configured to selectively evaluate a selection of documents from the one or more clusters, and apply one or more label corrections to the ANN;
- and
- the ANN configured to generate an update to the document collection incorporating the applied one or more label corrections.
A computer program product to utilize machine learning to facilitate document processing, the computer program product comprising:
- a tangible computer readable storage medium having program code embedded therewith, the program code executable by a processor to:
- cause an artificial neural network (ANN) to subject a document collection to table region identification within one or more discretized contiguous areas;
- subject the document collection to clustering, including:
- leverage the ANN to extract one or more structures and location of the one or more structures in the document collection;
- and
- assign the documents within the document collection to one or more clusters responsive to the leveraged ANN, wherein each cluster includes one or more documents having a content characteristic;
- selectively evaluate a selection of documents from the one or more clusters, and apply one or more label corrections from the evaluation to the ANN;
- and
- the ANN configured to generate an update to the document collection incorporating the applied one or more label corrections.
A computer-implemented method comprising:
- causing an artificial neural network (ANN) to subject a document collection to table region identification within one or more discretized contiguous areas;
- subjecting the document collection to clustering, including:
- leveraging the ANN to extract one or more structures and location of the one or more structures in the document collection;
- and
- assigning the documents within the document collection to one or more clusters responsive to the leveraged ANN, wherein each cluster includes one or more documents having a content characteristic;
- selectively evaluating a selection of documents from the one or more clusters, and applying one or more label corrections to the ANN;
- and
- the ANN generating an update to the document collection incorporating the applied one or more label corrections.
Dependencies are inferred from the claim text.