Data store organizing data using semantic classification

• US 9,613,125 B2
• Filed: 10/11/2012
• Issued: 04/04/2017
• Est. Priority Date: 10/11/2012
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Casella dos Santos, Mariana, Montyne, Frank
• Primary Examiner(s): Lu, Charles
• Application Number:
  US13/649,866
• Publication Number:
  US 20140108460A1
• Time in Patent Office: 1,636 Days
• Field of Search: 707741
• CPC Class Codes: G06F 16/284 Relational databases
• CPC Class Codes: G06F 16/35 Clustering; Classification

Abstract

Data stores that store content units and annotations regarding the content units derived through a semantic interpretation of the content units. When annotations are stored in a database, different parts of an annotation may be stored in different tables of the database. For example, one or more tables of the database may store all semantic classifications for the annotations, while one or more other tables may store content of all of the annotations. A user may be permitted to provide natural language queries for searching the database. A natural language query may be semantically interpreted to determine one or more annotations from the query. The semantic interpretation of the query may be performed using the same annotation model used to determine annotations stored in the database. Semantic classifications and format of the annotations for a query may be the same as one or more annotations stored in the database.

Claims

1. A method comprising:

storing in a database a first annotation and a second annotation, the first annotation relating to a first content unit and comprising a first semantic label and first content, the first semantic label comprising a term that does not appear in the first content and indicating a semantic classification of the first content, the second annotation relating to a second content unit and comprising a second semantic label and second content, the second semantic label comprising a term that does not appear in the second content and indicating a semantic classification of the second content, the semantic classification of the second content being different from the semantic classification of the first content, wherein the semantic classification of the first content indicates a meaning of the first content in context of the first content unit,

wherein the first content does not explicitly appear in the first content unit, the term that indicates the semantic classification of the first content indicating a meaning of the first content in context of the first content unit from which the first content was determined,

wherein the second content is a text excerpt of text of the second content unit,

wherein the semantic classification of the second content indicates that the second content is an organizational and/or grammatical element of the text of the second content unit,

wherein the storing comprises

storing the first semantic label for the first annotation and the second semantic label for the second annotation in a first table of the database, and

storing the first content of the first annotation and the second content of the second annotation in at least one second table of the database different from the first table.

2. The method of claim 1, further comprising:

storing in the database a plurality of annotations in addition to the first annotation and the second annotation, the plurality of annotations comprising a plurality of semantic labels that each indicates a semantic classification among a plurality of semantic classifications, at least some of the plurality of semantic classifications indicating a classification of meaning, each annotation of the plurality of annotations comprising a semantic label from among the plurality of semantic labels, the plurality of semantic classifications comprising the semantic classification of the first content and the semantic classification of the second content,

wherein the storing the plurality of annotations in the database comprises storing each of the plurality of semantic labels in the first table.

3. The method of claim 1, wherein the storing the first semantic label and the second semantic label in the first table of the database comprises storing the first semantic label and the second semantic label in one data structure in at least one computer-readable storage medium, the data structure specifying an organization of the first table.

4. The method of claim 1, wherein:

the at least one second table comprises a plurality of content tables, each content table of the plurality of content tables comprising at least one content field for storing content of an annotation, the plurality of content tables comprising a first content table and a second content table, the first content table comprising a first content field for storing annotation content of a first data type and the second content table comprising a second content field for storing annotation content of a second data type, the first data type being different from the second data type; and

the storing the first content in the at least one second table comprises;

determining a data type of the first content;

when it is determined that the first content is of the first data type, storing the first content in the first content table in response to determining that the first content is of the first data type; and

when it is determined that the first content is of the second data type, storing the first content in the second content table in response to determining that the first content is of the second data type.

5. The method of claim 4, wherein:

the method further comprises maintaining, for each one of the plurality of content tables, at least one index corresponding to the at least one content field of the content table, wherein the at least one index for the first content table comprises a first index corresponding to the first content field and the at least one index for the second content table comprises a second index corresponding to the second content field; and

the storing the first content in the at least one second table further comprises;

when it is determined that the first content is of the first data type, updating the first index on the first content field following storage of the first content in the first content table in response to determining that the first content is of the first data type; and

when it is determined that the first content is of the second data type, updating the second index on the second content field following storage of the first content in the second content table in response to determining that the first content is of the second data type.

6. The method of claim 1, further comprising:

storing, in the first table, a first content unit identifier identifying the first content unit to which the first annotation relates, the storing comprising identifying, in the first table, that the first content unit identifier is associated with the first annotation;

storing, in the first table, a second content unit identifier identifying the second content unit to which the second annotation relates, the storing comprising identifying, in the first table, that the second content unit identifier is associated with the second annotation; and

storing the first content unit and the second content unit in the database.

7. The method of claim 1, further comprising:

storing in the database a third annotation relating to the first content unit, the third annotation comprising a third semantic label and third content, the third semantic label indicating a semantic classification of the third content, wherein the storing comprises

storing the third semantic label for the third annotation in the first table of the database, and

storing the third content of the third annotation in the at least one second table of the database; and

associating the first annotation with the third annotation in the database, wherein the associating comprises storing, in a third table of the database, a first identifier for the first annotation and a third identifier for the third annotation.

8. The method of claim 7, wherein:

the first content comprises content explicitly recited in the first content unit;

the third content comprises content not explicitly recited in the first content unit; and

the third content is related to the first content and provides information regarding the first content.

9. The method of claim 1, wherein:

the first content unit records a clinical encounter; and

the storing the first annotation related to the first content unit comprises storing at least one fact that has a medical significance.

10. The method of claim 9, wherein the storing the at least one fact comprises storing at least one medical fact that is not explicitly recited in the first content unit.

11. The method of claim 1, wherein the first content unit and the second content unit are the same content unit.

12. The method of claim 2, wherein:

at least some of the plurality of semantic classifications indicate a classification of a role of content relative to a content unit from which the content was determined;

storing the plurality of annotations comprises storing a third annotation relating to the first content unit, the third annotation comprising a third semantic label and third content, the third semantic label indicating a role of the third content relative to the first content unit; and

wherein storing each of the plurality of semantic labels in the first table comprises storing the third semantic label in the first table.

13. The method of claim 12, wherein:

the third annotation is a structural annotation identifying a structural element appearing in text of the first content unit;

the third content of the third annotation identifies the structural element; and

the third semantic label identifies a structure type of the structural element.

14. The method of claim 2, wherein:

the at least one data store stores, separate from the first table and the second table, a set of all semantic classifications that are available for inclusion in annotations as semantic labels indicating semantic classification of content of annotations; and

the plurality of semantic classifications for the plurality of annotations are a subset of the set of semantic classifications.

15. The method of claim 1, wherein the semantic classification of the first content indicates a meaning of the first content in context of a portion of the first content unit from which the first content was determined.

16. At least one computer-readable storage medium having encoded thereon computer-executable instructions that, when executed by at least one computer, cause the at least one computer to carry out a method comprising:

storing in a database a first annotation and a second annotation, the first annotation and second annotation relating to a first content unit,

wherein the first annotation comprises a first semantic label and first content,

wherein the first semantic label comprises a term that does not appear in the first content and indicates a semantic classification of the first content,

wherein the first content was determined from an analysis of the first content unit and does not explicitly appear in the first content unit, the term that indicates the semantic classification of the first content indicating a meaning of the first content in context of the first content unit from which the first content was determined,

wherein the second semantic label comprises a term that does not appear in the second content and indicates a semantic classification of the second content,

wherein the second content is a text excerpt of text of the first content unit, the semantic classification of the second content being different from the semantic classification of the first content, wherein the term that indicates the semantic classification of the second content indicates that the second content is a grammatical and/or structural element of the text of the first content unit,

wherein the storing comprises

storing the first semantic label for the first annotation and the second semantic label for the second annotation in a first table of the database, and

storing the first content of the first annotation and the second content of the second annotation in at least one second table of the database different from the first table.

17. The at least one computer-readable storage medium of claim 16, wherein the storing the first semantic label and the second semantic label in the first table of the database comprises storing the first semantic label and the second semantic label in one data structure in at least one computer-readable storage medium, the data structure specifying an organization of the first table.

18. The at least one computer-readable storage medium of claim 16, wherein:

the at least one second table comprises a plurality of content tables, each content table of the plurality of content tables comprising at least one content field for storing content of an annotation, the plurality of content tables comprising a first content table and a second content table, the first content table comprising a first content field for storing annotation content of a first data type and the second content table comprising a second content field for storing annotation content of a second data type, the first data type being different from the second data type; and

the storing the first content in the at least one second table comprises;

determining a data type of the first content;

when it is determined that the first content is of the first data type, storing the first content in the first content table in response to determining that the first content is of the first data type; and

when it is determined that the first content is of the second data type, storing the first content in the second content table in response to determining that the first content is of the second data type.

19. An apparatus comprising:

at least one processor; and

at least one computer-readable storage medium having encoded thereon executable instructions that, when executed by the at least one processor, cause the at least one processor to carry out a method of interpreting content units using an annotation model, wherein the annotation model defines a plurality of annotation types that may be determined from content units, wherein the annotation model defines, for each annotation type of the plurality of annotation types, a semantic label for the annotation type and a type of content for the annotation type, the semantic label for each annotation type of the plurality indicating a semantic classification of the type of content for that annotation type, wherein the annotation model identifies, for the plurality of annotation types, a plurality of types of content that are to be included in the plurality of annotation types, the method comprising;

determining, based at least in part on the annotation model that defines the plurality of annotation types, a first annotation for a first content unit from the first content unit and a second annotation for a second content unit from the second content unit, the first annotation being of a first annotation type defined by the annotation model and the second annotation being of a second annotation type defined by the annotation model,

wherein determining the first annotation comprises determining from an analysis of a portion of text of the first content unit that the first annotation type applies to the portion of the text of the first content unit,

wherein determining that the first annotation type applies to the portion of the text of the first content unit comprises

determining, from the portion of the text, content of the first annotation that is of the type of content defined by the annotation model for the first annotation type and is semantically classified in context of the first content unit according to the semantic classification indicated by the semantic label defined by the annotation model for the first annotation type,

wherein in the first annotation, the first semantic label defined by the annotation model for the first annotation type comprises a term that does not appear in the first content and indicates a semantic classification of the first content, the semantic classification of the first content indicating a meaning of the first content in context of the first content unit,

wherein the second annotation relates to the second content unit and comprises a second semantic label and second content, the second semantic label comprising a term that does not appear in the second content and indicates a semantic classification of the second content, the semantic classification of the second content being different from the semantic classification of the first content, the semantic classification of the second content indicating a meaning of the second content in context of the second content unit; and

storing the first annotation and the second annotation in a database, wherein the storing comprises

storing the first semantic label for the first annotation and the second semantic label for the second annotation in a first table of the database, and

storing the first content of the first annotation and the second content of the second annotation in at least one second table of the database different from the first table.

20. The apparatus of claim 19, wherein:

the at least one second table comprises a plurality of content tables, each content table of the plurality of content tables comprising at least one content field for storing content of an annotation, the plurality of content tables comprising a first content table and a second content table, the first content table comprising a first content field for storing annotation content of a first data type and the second content table comprising a second content field for storing annotation content of a second data type, the first data type being different from the second data type; and

the storing the first content in the at least one second table comprises;

determining a data type of the first content;

when it is determined that the first content is of the first data type, storing the first content in the first content table in response to determining that the first content is of the first data type; and

when it is determined that the first content is of the second data type, storing the first content in the second content table in response to determining that the first content is of the second data type.