Utilizing classification and text analytics for annotating documents to allow quick scanning

• US 10,509,852 B2
• Filed: 03/27/2014
• Issued: 12/17/2019
• Est. Priority Date: 12/10/2012
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Emanuel, Barton W., Paulis, Mark W., Roboff, Mark L.
• Primary Examiner(s): Zuberi, Mohammed H
• Application Number:
  US14/227,960
• Publication Number:
  US 20140215304A1
• Time in Patent Office: 2,091 Days
• Field of Search: 704 10, 704 1, 704 7, 704 9, 707708, 715854
• CPC Class Codes: G06F 40/169 Annotation, e.g. comment data or footnotes
• CPC Class Codes: G06F 40/211 Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
• CPC Class Codes: G06F 40/253 Grammatical analysis; Style critique
• CPC Class Codes: G06F 40/268 Morphological analysis
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

Classification, text analytics, and natural language processing are used to evaluate passages, extract text, identify concepts, and provide visual cues and notations to assist readers in scanning and evaluating large amounts of information in a document.

Claims

1. A computer-implemented method for annotating a document comprising:

determining a type of the document using a classifier, wherein the type of document is a format within which the document is written;

determining a subject domain of the document using the classifier, wherein the subject domain is a field, topic or genre of content set forth in the document;

segmenting the document into one or more paragraphs and sections based on the document'"'"'s structure using the classifier;

determining and driving an annotation strategy based on the type of document using information from an annotation model, wherein the annotation strategy specifies one or more annotations to utilize and a location in the document'"'"'s original text and on a margin of the document for the one or more annotations;

loading a domain model based on the subject domain into a text analytics system, wherein the subject domain determines which domain model to load into the text analytics system and the domain model identifies terms, phrases, entities, and concepts of the subject domain to be annotated in the document;

providing the one or more annotations in the document'"'"'s original text and on the margin of the document for the paragraphs of the document based on the domain model and the annotation model using the text analytics system, wherein the one or more annotations comprise domain-specific keywords and concepts;

aggregating the one or more annotations in the margin of the document for the paragraphs of the document into one or more section-level aggregated annotations for the sections of the document based on the annotation model using the text analytics system;

annotating the document with the one or more annotations in the document'"'"'s original text and on the margin of the document for the paragraphs, and with the one or more section-level aggregated annotations for the sections of the document, using a custom viewer/renderer application; and

rendering the document including the one or more annotations and the one or more section-level aggregated annotations using the custom viewer/renderer application.

2. The method of claim 1, wherein the annotations for the paragraphs of the document are selected from the group consisting of:

parts of speech, phrases, proper names, dates, and relationships.

3. The method of claim 1, further comprising normalizing the one or more annotations for the paragraphs of the document by:

identifying sets of one or more synonyms; and

determining normal forms using a lexical ontology.

4. The method of claim 1, further comprising:

removing stop words; and

removing words frequently occurring throughout the document, the subject domain, or a set of documents of which the document is a member.

5. The method of claim 1, wherein the annotating comprises annotating the document'"'"'s original text by highlighting.

6. The method of claim 1, further comprising saving the rendering to disk.

7. The method of claim 1, further comprising generating an extensible markup language (XML) file of the one or more annotations, wherein a renderer is configured to execute in standalone mode based on the XML, file without having to access a network to perform classification or text analytics.

8. The method of claim 1, wherein the one or more annotations to utilize is selected from the group consisting of:

word, phrase, concept cloud, list, quality indicator, and timeline.