SUGGESTING RELEVANT TERMS DURING TEXT ENTRY

• US 20130212475A1
• Filed: 10/31/2011
• Published: 08/15/2013
• Est. Priority Date: 11/01/2010
• Status: Active Grant

• Current Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Original Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Inventors: Cohen-Solal, Eric, Lee, Michael Chun-Chieh, Qian, Yuechen
• Granted Patent: US 9,886,427 B2
• Time in Patent Office: Days
• US Class Current: 715/261
• CPC Class Codes: G06F 40/166 Editing, e.g. inserting or deleting
• CPC Class Codes: G06F 40/274 Converting codes to words; Guess-ahead of partial word inputs
• CPC Class Codes: G06F 40/284 Lexical analysis, e.g. tokenisation or collocates

Abstract

A system for suggesting autocompletion terms during text entry of a report. A text entry unit (2) for enabling a user to enter a text into a current report (1). A term selector (11) for selecting at least one frequently co-occurring term (12), based on an extracted term (8), the extracted term section (9), the current section (10), and a co-occurrence statistic (7). An indicator (13) provides an indication of the at least one frequently co-occurring term (12) to the user. A system for analyzing reports comprises a co-occurrence statistics generator (26) for generating a plurality of co-occurrence statistics, a co-occurrence statistic being indicative of a first term, a first section, a second term, a second section, and a frequency in which reports contain the first term in the first section in combination with the second term in the second section.

Claims

1. A system for suggesting autocompletion terms during text entry of a report, comprising

a text entry unit for enabling a user to enter a text into a current report (1);

a section analyzer for determining a plurality of sections of the current report;

a current section detector for detecting a section of the current report which the user is working on, thus obtaining a current section;

a term extractor for extracting a term occurring in the current report, thus obtaining an extracted term, and identifying a section of the current report in which the extracted term occurs, thus obtaining an extracted term section, wherein the extracted term section and the current section are different sections;

a co-occurrence accessing unit for accessing a plurality of co-occurrence statistics, a co-occurrence statistic being indicative of at least one first term, at least one first section, a second term, a second section, and a frequency in which reports in a knowledge domain contain the at least one first term in the at least one first section in combination with the second term in the second section;

a term selector for selecting at least one frequently co-occurring term, based on the extracted term, the extracted term section, the current section, and at least one of the co-occurrence statistics, wherein the term selector selects the at least one frequency co-occurring term for a portion of a term entered in the current section based on at least one of the co-occurrence statistics associated with the extracted term in the extracted term section;

an indicator for providing an indication of the at least one frequently co-occurring term to the user.

3. The system according to claim 1, wherein a report comprises a document and wherein a section comprises a section heading and a section body.

4. The system according to claim 1, wherein the report comprises a plurality of files, and wherein different files comprise different sections of the report.

5. The system according to claim 1, wherein the extracted term and/or the first term comprise an expression comprising a plurality of words.

6. The system according to claim 1, comprising a natural language processor for associating the extracted term and/or the at least one first term with at least one respective ontological concept in an ontology, and wherein the co-occurrence statistic relates to the probability of the co-occurrence of the at least one ontological concept with the second term.

7. The system according to claim 1, wherein the term selector is operatively coupled to the text entry unit and arranged for receiving the portion of the term which is entered by the user, and arranged for selecting the frequently co-occurring term, based on the received portion of the term.

8. The system according to claim 7, wherein the term selector is arranged for selecting at least one frequently co-occurring term of which the beginning matches the received portion of the term.

9. The system according to claim 1, wherein the reports include medical patient reports and the sections include a patient history section, a clinical findings section, and/or a diagnosis section.

10. A workstation comprising the system according to either or both of claims 1.

2. A system for analyzing reports, comprising

a section analyzer for determining a division of a plurality of reports into sections;

a common section determiner for determining a plurality of sections common to a plurality of the reports;

a term extractor for extracting a plurality of terms from the common sections of the reports, and associating each term with the section and report in which it occurs;

a co-occurrence statistics generator for generating a plurality of co-occurrence statistics, a co-occurrence statistic being indicative of at least one first term, at least one first section, a second term, a second section, and a frequency in which reports contain the at least one first term in the at least one first section in combination with the second term in the second section, wherein the at least one first section is different from the second section.

11. A method of suggesting autocompletion terms during text entry of a report, comprising

enabling a user to enter a text into a current report;

determining a plurality of sections of the current report;

detecting a section of the current report which the user is working on, thus obtaining a current section;

extracting a term occurring in the current report, thus obtaining an extracted term, and identifying a section of the current report in which the extracted term occurs, thus obtaining an extracted term section, wherein the extracted term section and the current section are different sections;

accessing a plurality of co-occurrence statistics, a co-occurrence statistic being indicative of at least one first term, at least one first section, a second term, a second section, and a frequency in which reports in a knowledge domain contain the at least one first term in the at least one first section in combination with the second term in the second section;

selecting at least one frequently co-occurring term, based on the extracted term, the extracted term section, the current section, and at least one of the co-occurrence statistics, wherein the selection of the at least one frequency co-occurring term for a portion of a term entered in the current section is based on at least one of the co-occurrence statistics associated with the extracted term in the extracted term section; and

providing an indication of the at least one frequently co-occurring term to the user.

13. A computer program product comprising instructions for causing a processor system to perform the method according to claim 11.

12. A method of analyzing a plurality of reports, comprising

determining a division of the reports into sections;

determining a plurality of sections common to a plurality of the reports;

extracting a plurality of terms from the common sections of the reports, and associating each term with the section and report in which it occurs; and

generating a plurality of co-occurrence statistics, a co-occurrence statistic being indicative of at least one first term, at least one first section, a second term, a second section, and a frequency in which reports contain the at least one first term in the at least one first section in combination with the second term in the second section, wherein the at least one first section is different from the second section.