Interactive visual representation of semantically related extracted data
Techniques are disclosed to provide an interactive visual representation of semantically related extracted data. In various embodiments, a plurality of data entities are extracted from a file, each entity comprising a key-value pair. One or more sets of related entities, each set comprising an occurrence of a defined repeating type of entity set, are identified among the plurality of data entities. Data associating the one or more sets of related entities with the file and the defined repeating type of entity set are stored.
1. A system, comprising:
a communication interface configured to receive a file; and
a processor coupled to the communication interface and configured to:
extract a plurality of data entities from the file, each entity comprising a key-value pair;
identify among the plurality of data entities one or more sets of related entities, each set comprising an occurrence of a defined repeating type of entity set; and
store data associating the one or more sets of related entities with the file and the defined repeating type of entity set.
2. The system of claim 1 , wherein the processor is further configured to use the stored data associating the one or more sets of related entities with the file and the defined repeating type of entity set to provide an extracted data validation user interface in which each of the one or more sets of related entities is displayed as a corresponding row.
3. The system of claim 2 , wherein the extracted data validation user interface displays data extracted from a plurality of files and for each file the extracted data validation user interface includes a display of one or more rows, each row corresponding to and displaying extracted values associated with a corresponding set of related entities.
4. The system of claim 3 , wherein extracted data validation user interface provides one or more features associated with a spreadsheet application.
5. The system of claim 3 , wherein the extracted data validation user interface includes for at least a subset of entities displayed in at least a subset of rows a visual indication of a state of validation of one or both of a data value and a label associated with that entity.
6. The system of claim 5 , wherein the visual indication includes a color code reflecting a degree of confidence.
7. The system of claim 5 , wherein the extracted data validation user interface includes for each of at least said subset of entities displayed in at least said subset of rows a user-selectable control to access a file data validation interface for one or both of the entity and a corresponding one of said plurality of files with which the entity is associated.
8. The system of claim 7 , wherein the file data validation interface includes a first region in which data representing the corresponding one of said plurality of files is displayed and a second region that includes an interactive data validation element to validate one or both of a data value and a label associated with an entity comprising the file.
9. The system of claim 8 , wherein said interactive data validation element is displayed in response to selection of a representation of said entity within the first region.
10. The system of claim 8 , wherein the processor is configured to receive via the interactive data validation element a user input indicating validation of one or both of the data value and the label associated with the entity; and the processor is further configured to update the extracted data validation user interface to reflect validation of the entity.
11. The system of claim 10 , wherein the file comprises a first file included in said plurality of files and the validated entity comprises a first entity; and wherein the processor is configured to update the extracted data validation user interface to reflect validation of the first entity; and to update a second entity associated with a second file based on a reevaluation in real time of a confidence associated with a one or both of a predicted value and a predicted label associated with the second entity, the reevaluation being determined based at least in part on the validation of the first entity.
12. The system of claim 10 , wherein the processor is further configured to update a data extraction model based at least in part on the validation of the entity.
13. The system of claim 12 , wherein the processor is configured to update the data extraction model at least in part by providing data associated with validation of the entity to a machine learning process or module.
14. The system of claim 1 , wherein the processor is further configured to receive a definition of said repeating type of entity set.
15. The system of claim 14 , wherein the definition is received via a data extraction configuration user interface and wherein the processor is further configured to display the data extraction configuration user interface.
16. The system of claim 15 , wherein the data extraction configuration user interface enables a user to select an instance of a set of related entities as represented in a displayed representation of a file with which the set of related entities is associated and to define the repeating type of entity set in relation to the selected instance of the set of related entities.
17. A method, comprising:
extracting a plurality of data entities from a file, each entity comprising a key-value pair;
identifying among the plurality of data entities one or more sets of related entities, each set comprising an occurrence of a defined repeating type of entity set; and
storing data associating the one or more sets of related entities with the file and the defined repeating type of entity set.
18. The method of claim 17 , further comprising using the stored data associating the one or more sets of related entities with the file and the defined repeating type of entity set to provide an extracted data validation user interface in which each of the one or more sets of related entities is displayed as a corresponding row.
19. The method of claim 18 , wherein the extracted data validation user interface displays data is extracted from a plurality of files and for each file the extracted data validation user interface includes a display of one or more rows, each row corresponding to and displaying extracted values associated with a corresponding set of related entities.
20. A computer program product embodied in a non-transitory computer readable medium, comprising computer instructions for:
extracting a plurality of data entities from a file, each entity comprising a key-value pair;
identifying among the plurality of data entities one or more sets of related entities, each set comprising an occurrence of a defined repeating type of entity set; and
storing data associating the one or more sets of related entities with the file and the defined repeating type of entity set.