Data comparison using natural language processing models

1 . A method comprising:

receiving a first set of labeled data from a first data source;

receiving a text string from a second data source corresponding to one or more particular fields of a document;

generating a second set of labeled data based on parsing of the text string, the generating comprising:

determining, from the second set of labeled data, first text portions associated with vesting information, wherein determining the first text portions associated with the vesting information comprises comparing the particular text portions with entries in a domain knowledge base corresponding to particular vesting types; and

generating the second set of labeled data that includes labeled vesting information determined from the text string;

performing a comparison between the first set of labeled data and the second set of labeled data including comparing the labeled vesting information with a current vesting string in the first set of labeled data;

determining whether the comparison results satisfy a user adjustable threshold level indicating how strictly one or more comparison rules are applied when performing the comparison; and

generating an output based on the comparison and the threshold.

2 . The method of claim 1 , wherein prior to performing the comparison, performing one or more pre-comparison checks, the pre-comparison checks comprising a check to determine whether the first or second set of labeled data indicates vesting information that requires manual evaluation.

3 . The method of claim 1 , wherein the document is a real estate transaction document and wherein the second set of labeled data includes labels for particular fields in the document including labels associated with vesting and marital status.

4 . The method of claim 3 , further comprising comparing a labeled vesting type from the second set of labeled data with a labeled marital status for consistency between vesting type and marital status text in the document.

5 . The method of claim 3 , wherein performing the comparison comprises comparing a vesting type from the first set of labeled data with a vesting type from the second set of labeled data, the comparing including using the domain knowledge base to determine whether two different vesting types are equivalent.

6 . The method of claim 3 , wherein performing the comparison comprises comparing a vesting type from the first set of labeled data with a vesting type from the second set of labeled data, the comparing including determining whether non-equivalent vesting types are considered consistent.

7 . The method of claim 1 , wherein generating the output comprises determining that the vesting information is consistent such that a new vesting deed is not needed.

8 . A system comprising:

one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:

receiving a first set of labeled data from a first data source;

receiving a text string from a second data source corresponding to one or more particular fields of a document;

generating a second set of labeled data based on parsing of the text string, the generating comprising:

determining, from the second set of labeled data, first text portions associated with vesting information, wherein determining the first text portions associated with the vesting information comprises comparing the particular text portions with entries in a domain knowledge base corresponding to particular vesting types; and

generating the second set of labeled data that includes labeled vesting information determined from the text string;

performing a comparison between the first set of labeled data and the second set of labeled data including comparing the labeled vesting information with a current vesting string in the first set of labeled data;

determining whether the comparison results satisfy a user adjustable threshold level indicating how strictly one or more comparison rules are applied when performing the comparison; and

generating an output based on the comparison and the threshold.

9 . The system of claim 8 , wherein prior to performing the comparison, performing one or more pre-comparison checks, the pre-comparison checks comprising a check to determine whether the first or second set of labeled data indicates vesting information that requires manual evaluation.

10 . The system of claim 8 , wherein the document is a real estate transaction document and wherein the second set of labeled data includes labels for particular fields in the document including labels associated with vesting and marital status.

11 . The system of claim 10 , the operations further comprising comparing a labeled vesting type from the second set of labeled data with a labeled marital status for consistency between vesting type and marital status text in the document.

12 . The system of claim 10 , wherein performing the comparison comprises comparing a vesting type from the first set of labeled data with a vesting type from the second set of labeled data, the comparing including using the domain knowledge base to determine whether two different vesting types are equivalent.

13 . The system of claim 10 , wherein performing the comparison comprises comparing a vesting type from the first set of labeled data with a vesting type from the second set of labeled data, the comparing including determining whether non-equivalent vesting types are considered consistent.

14 . The system of claim 8 , wherein generating the output comprises determining that the vesting information is consistent such that a new vesting deed is not needed.

15 . One or more non-transitory computer storage media encoded with computer program instructions that when executed by one or more computers cause the one or more computers to perform operations comprising:

receiving a first set of labeled data from a first data source;

receiving a text string from a second data source corresponding to one or more particular fields of a document;

generating a second set of labeled data based on parsing of the text string, the generating comprising:

determining, from the second set of labeled data, first text portions associated with vesting information, wherein determining the first text portions associated with the vesting information comprises comparing the particular text portions with entries in a domain knowledge base corresponding to particular vesting types; and

generating the second set of labeled data that includes labeled vesting information determined from the text string;

performing a comparison between the first set of labeled data and the second set of labeled data including comparing the labeled vesting information with a current vesting string in the first set of labeled data;

determining whether the comparison results satisfy a user adjustable threshold level indicating how strictly one or more comparison rules are applied when performing the comparison; and

generating an output based on the comparison and the threshold.

16 . The computer storage media of claim 15 , wherein prior to performing the comparison, performing one or more pre-comparison checks, the pre-comparison checks comprising a check to determine whether the first or second set of labeled data indicates vesting information that requires manual evaluation.

17 . The computer storage media of claim 15 , wherein the document is a real estate transaction document and wherein the second set of labeled data includes labels for particular fields in the document including labels associated with vesting and marital status.

18 . The computer storage media of claim 17 , further comprising comparing a labeled vesting type from the second set of labeled data with a labeled marital status for consistency between vesting type and marital status text in the document.

19 . The computer storage media of claim 17 , wherein performing the comparison comprises comparing a vesting type from the first set of labeled data with a vesting type from the second set of labeled data, the comparing including using the domain knowledge base to determine whether two different vesting types are equivalent.

20 . The computer storage media of claim 17 , wherein performing the comparison comprises comparing a vesting type from the first set of labeled data with a vesting type from the second set of labeled data, the comparing including determining whether non-equivalent vesting types are considered consistent.

21 . The computer storage media of claim 15 , wherein generating the output comprises determining that the vesting information is consistent such that a new vesting deed is not needed.