Error correction for tables in document conversion
A method is presented for error correction of tabular data in document conversion. The method includes identifying errors from tabular data transformation by employing an error/invalidation checking module and correcting the identified errors from the tabular data transformation by employing an error correction module. The error correction module includes identifying a main structure pattern from common row structures, concatenating separate keywords according to natural language processing models employing training data obtained from a plurality of candidate tabular data, adjusting cells in the tabular data based on a domain-specific knowledge database including the training data in combination with linguistic and semantic knowledge, merging partial tabular data pieces, and generating an adjusted table as output on a display of a computing device.
1. A method for error correction of tabular data in document conversion, the method comprising: identifying errors from tabular data transformation by employing an error/invalidation checking module; and correcting the identified errors from the tabular data transformation by employing an error correction module, the error correction module comprising: identifying a main structure pattern from common row structures; concatenating separate keywords according to natural language processing models employing training data obtained from a plurality of candidate tabular data; adjusting cells in the tabular data based on a domain-specific knowledge database including the training data in combination with linguistic and semantic knowledge; merging partial tabular data pieces; and generating an adjusted table as output on a display of a computing device.
2. The method of claim 1 , wherein the identified errors include at least meaningless cell text strings, incorrectly split text strings, incorrectly combined text strings, cell value inconsistencies, and structure inconsistencies.
3. The method of claim 1 , wherein the error/invalidation checking module includes a structure invalidation checking component, a text invalidation checking component, and a value invalidation checking component.
4. The method of claim 1 , wherein the error correction module includes a local correction proposer and a global correction decider.
5. The method of claim 1 , wherein the adjusted table is further evaluated based on the following criteria: a weighted likelihood of remaining possible inconsistencies in the adjusted table and a number of changes needed to produce the adjusted table.
6. The method of claim 5 , wherein the adjusted table is converted to a final table by employing a weighted sum of the criteria.
7. The method of claim 6 , wherein, when the weighted sum is above a predetermined threshold, the final table is flagged to receive user input.
8. A non-transitory computer-readable storage medium comprising a computer-readable program executed on a processor in a data processing system for error correction of tabular data in document conversion, wherein the computer-readable program when executed on the processor causes a computer to perform the steps of: identifying errors from tabular data transformation by employing an error/invalidation checking module; and correcting the identified errors from the tabular data transformation by employing an error correction module, the error correction module comprising: identifying a main structure pattern from common row structures; concatenating separate keywords according to natural language processing models employing training data obtained from a plurality of candidate tabular data; adjusting cells in the tabular data based on a domain-specific knowledge database including the training data in combination with linguistic and semantic knowledge; merging partial tabular data pieces; and generating an adjusted table as output on a display of a computing device.
9. The non-transitory computer-readable storage medium of claim 8 , wherein the identified errors include at least meaningless cell text strings, incorrectly split text strings, incorrectly combined text strings, cell value inconsistencies, and structure inconsistencies.
10. The non-transitory computer-readable storage medium of claim 8 , wherein the error/invalidation checking module includes a structure invalidation checking component, a text invalidation checking component, and a value invalidation checking component.
11. The non-transitory computer-readable storage medium of claim 8 , wherein the error correction module includes a local correction proposer and a global correction decider.
12. The non-transitory computer-readable storage medium of claim 8 , wherein the adjusted table is further evaluated based on the following criteria: a weighted likelihood of remaining possible inconsistencies in the adjusted table and a number of changes needed to produce the adjusted table.
13. The non-transitory computer-readable storage medium of claim 12 , wherein the adjusted table is converted to a final table by employing a weighted sum of the criteria.
14. The non-transitory computer-readable storage medium of claim 13 , wherein, when the weighted sum is above a predetermined threshold, the final table is flagged to receive user input.
15. An apparatus for error correction of tabular data in document conversion, the apparatus comprising:
an error/invalidation checking module to identify errors from tabular data transformation; and
an error correction module to correct the identified errors from the tabular data transformation, the error correction module configured to:
identify a main structure pattern from common row structures;
concatenate separate keywords according to natural language processing models employing training data obtained from a plurality of candidate tabular data;
adjust cells in the tabular data based on a domain-specific knowledge database including the training data in combination with linguistic and semantic knowledge;
merge partial tabular data pieces; and
generate an adjusted table as output on a display of a computing device.
16. The apparatus of claim 15 , wherein the identified errors include at least meaningless cell text strings, incorrectly split text strings, incorrectly combined text strings, cell value inconsistencies, and structure inconsistencies.
17. The apparatus of claim 15 , wherein the error/invalidation checking module includes a structure invalidation checking component, a text invalidation checking component, and a value invalidation checking component.
18. The apparatus of claim 15 , wherein the error correction module includes a local correction proposer and a global correction decider.
19. The apparatus of claim 15 , wherein the adjusted table is further evaluated based on the following criteria: a weighted likelihood of remaining possible inconsistencies in the adjusted table and a number of changes needed to produce the adjusted table.
20. The apparatus of claim 18 , wherein the adjusted table is converted to a final table by employing a weighted sum of the criteria.