2.90:Confidence Multiplier and Output Confidence (Property)

About

Weighted rules is an informal title given to the practical application of the Confidence Multiplier property of a Data Type. Its practical application allows a user to arbitrarily set the confidence of a result of a particular Data Type in order to allow that Data Type to appear more (or less) favorable to a parent Data Type that is leveraging the Order By property configured to the Confidence setting.

Use Cases

Weighted Rules can be used in cases where one is trying to find an element of data which can appear on many similar types of forms that do not have a consistent method to identify where the data is.
For Example, on different forms, the best method to pick up a piece of data may be a Key-Value Pair, a Field Class, a simple pattern match, a pattern match leveraging FuzzyRegEx, or some other method.
One of the more recent methodologies for incorporating multiple extractors to be used by a single field has been coloquially referred to as Waterfall Extraction. This is done by organizing myriad extractors (and their numerous configurations) under a parent Data Type. The Order By property of the parent Data Type can then be set to the following: Position, Frequency, Confidence, Extractor, Length, Value.
Setting Order By to Confidence may be an interesting way to organize results, but typically, properly configured extractors always return their results at 100%. The confidence of a returned result has, historically, only been affected in one of two ways:

1. Data Type's (or a child Data Format's) regular expression pattern leverages FuzzyRegEx and it, as a result, had to insert, delete, or swap a character to match the pattern, thus generating a result less than 100% confident
2. Field Classes, by design leverage trained/weighted features and should not return results at 100% confidence
   

Considering this, a properly configured extractor can, and does, return results below 100%, and thus breaks the logical approach of organizing results by confidence. To elaborate, a result returned at 90% confidence could be more desirable than one returned at 100%.
Let's explore how and why.

How To

Here we'll explore a use case using a mortgage document.

OCR MisreadChild Data Type SetupWaterfall ExtractorDefault OutputGetting the Desired Result

In this example, an OCR error produced a misread the words “final loan” by not recognizing the space between them.

Three Data Types were established to find variations of a result.

FinalLoan	Final Loan	Fuzzy: Final Loan
A Data Type which uses a regular expression looking for the expression “finalloan” with no spaces.	A Data Type which uses a regular expression looking for the expression “final loan” with the space.	A Data Type which uses a fuzzy regular expression looking for the expression “final loan” with the space.

The Waterfall Extractor is a Data Type that is a parent or references all of the unique extractors for a piece of data and then determines which one should be given as a final output to a Data Field.

Using Order By set to Confidence and Direction set to Descending as the sort criteria, two extractors match with the highest confidence result given first. The FinalLoan extractor matched because it found “finalloan” with no spaces and it is not leveraging FuzzyRegEx, so it matched at 100%. The Final Loan extractor did not match, because it is not using FuzzyRegEx and it did not find a space between the two words so it did not consider it a match. The Fuzzy: Final Loan, leveraging FuzzyRegEx, matched because it was able to make the word “finalloan” into “final loan” by inserting a space and so it was a 90% match.

We would like the actual correct result of final loan to win. There are two ways to do this. One way would be to bump up the confidence of the fuzzy regular expression Data Type Fuzzy: Final Loan. This is done by modifying the Confidence Multiplier property in the Result Options of the Data Type' Fuzzy: Final Loan.

That works for this case, but what if there was another document where the OCR read the space between the two words correctly. In that case, the result from the Data Type Final Loan would match at 100%, and the Data Type Fuzzy: Final Loan, with the Confidence Multiplier property set to 1.2 would match at 120%. While this would technically yield the correct result, it is generally best practice to have the exact match return the highest percentage. There are a couple of ways to tackle this situation. One way would be to bump up the Confidence Multiplier property on the Data Type Final Loan to something like 1.3 But another way, would be to reduce the Confidence Multiplier property on the FinalLoan Data Type so that it returns less than 90%.

Let's change some settings to set this extractor up to return the results in the desired way; that being with the most right result weighted the highest.

Reset the Confidence Multiplier property in the Result Options property window for the Data Type Fuzzy: Final Loan.

Set the Confidence Multiplier' property in the Result Options property window for the Data Type FinalLoan to 0.75. The results on the parent Data Type will now show the un-weighted Data Type Fuzzy: Final Loan at a confidence of 90% (again, because a space was inserted), and the Data Type FinalLoan'l will show 75%. In the event another document is OCRed correctly with a space between the words, the Data Type Final Loan would return the exact match at 100%. The Data Type Fuzzy: Final Loan would also return 100% because the expression matched 100% with no substitutions. In order to make the exact matched always preferred, it would also be possible to set the Data Type Fuzzy: Final Loan Confidence Multiplier property to 0.99. But since both the fuzzy and the exact non-fuzzy Data Type matched 100%, it doesn’t really matter which one returns the result.