Incorporating content analytics and natural language processing into internet web browsers

• US 9,760,555 B2
• Filed: 04/21/2015
• Issued: 09/12/2017
• Est. Priority Date: 10/31/2014
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Fink, Patrick W., McNeil, Kristin E., Parker, Philip E., Werts, David B.
• Primary Examiner(s): Cortes, Howard
• Application Number:
  US14/691,808
• Publication Number:
  US 20160124928A1
• Time in Patent Office: 875 Days
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 40/169 Annotation, e.g. comment data or footnotes
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G16H 70/60 relating to pathologies

Abstract

Techniques disclosed herein describe generating annotations for text content rendered on a web browser. The web browser is configured with a content analytics plug-in configured to receive content to render in the web browser. The content itself includes unstructured text. The content analytics plug-in evaluates the unstructured text using content analytics and natural language processing (NLP) techniques to characterize the meaning of one or more passages in the unstructured text. The content analytics plug-in then generates annotations based on the characterizations to display when the web browser renders the content.

Claims

1. A method for generating annotations for web content rendered by a web browser, the method comprising:

receiving, via a plug-in executing in the web browser, input specifying a first annotator of an Unstructured Information Management Architecture (UIMA) that provides a plurality of processing stages, wherein the first annotator is of a plurality of annotators, wherein each annotator is associated with a respective distinct content domain and is configured to, in one of the processing stages, generate annotations associated with the associated distinct content domain for text passages;

receiving web content to render by the web browser, wherein the web content includes unstructured text;

evaluating by the first annotator, the unstructured text in one or more of the plurality of processing stages to identify one or more passages in the unstructured text determined to have a predicted meaning within the distinct content domain associated with the first annotator;

generating, in one or more of the processing stages by the first annotator, annotations for the identified one or more passages within the distinct content domain associated with the first annotator;

generating visual indicators for the annotations for the identified one or more passages within the distinct content domain associated with the first annotator; and

modifying the received web content to include the received web content, wherein the visual indicators are displayed when the browser renders the received web content.

2. The method of claim 1, wherein the annotations for each distinct content domain are rendered in a distinct visual indicator.

3. The method of claim 2, wherein the distinct content domains are each based on medical conditions.

4. The method of claim 1, further comprising, rendering the content and the annotations in the web browser.

5. The method of claim 4, further comprising:

during the rendering of the content and the annotations in the web browser, marking the one or more passages of the content that correspond to each of the annotations.

6. The method of claim 5, further comprising, upon receiving an indication that one of the marked passages is accessed in the web browser, generating a presentation of the corresponding annotation to display in the web browser.