System and method for identifying comparables

Patent metadata

Field	Value
Publication number	US-11720842-B2
Application number	US-201916731505-A
Country	US
Kind code	B2
Filing date	Dec 31, 2019
Priority date	Dec 31, 2019
Publication date	Aug 8, 2023
Grant date	Aug 8, 2023

Abstract

The invention relates to a computer implemented system and method for identification of comparables. The method may comprise: receiving input data from a plurality of data sources for a comparable, generating labeled training data for a function classifier by labeling historical search results for comparables, generating probabilistic training data for the primary product and service classifiers, training the primary product and service classifiers using the labeled training data and the probabilistic training data, determining the functions, products, services, and risks of the comparable using the corresponding classifiers, receiving attributes of a tested party, applying a scoring algorithm to calculate a similarity score for the comparable, generating a recommendation to accept the comparable, reject the comparable, or give additional scrutiny to determine acceptability, and automatically providing a written justification for the decision to accept or to reject the comparable.

First claim

What is claimed is:

1. A computer-implemented method for identifying comparables, the method comprising:
implementing a machine learning algorithm operating on a computer processor configured to:
collect, through an electronic interface, input data from a plurality of electronic data sources of a potential comparable for inclusion into a transfer pricing benchmarking set, wherein the transfer pricing benchmarking set is generated by a transfer pricing benchmarking activity which comprises search and selection of unrelated parties to establish a benchmark for the pricing of a cross-border transaction between one or more related parties under common ownership or control, wherein the electronic data sources include: a business description from a commercially available database, financial data of the potential comparable, and a SIC or NACE code associated with the potential comparable;
determine, via the machine learning algorithm, a function, product, and service of the potential comparable using a corresponding function classifier, product classifier, and service classifier;
receive, through the electronics interface, attributes of a tested party;
automatically execute, via the machine learning algorithm, a scoring process to calculate a similarity score for the potential comparable, wherein the similarity score represents a similarity between the potential comparable and the tested party;
automatically generate, via the machine learning algorithm, a recommendation to accept the potential comparable as an acceptable comparable for transfer pricing benchmarking, reject the potential comparable as a rejected comparable, or subject the potential comparable to further review; and
generate, via the machine learning algorithm, synthetic training data for the machine learning algorithm by (1) predicting a plurality of labels and keyword counts from the plurality of electronic data sources, (2) generating heuristic labels for the predicted plurality of labels based on a plurality of heuristic rules, the rules based on experience with prior similar problems, and (3) generating probabilistic training labels for a semi-supervised deep learning model; and
improve the machine learning algorithm by using the synthetic training data and feedback from the automatically generated recommendation to train the machine learning algorithm.

2. The computer-implemented method of claim 1 , further comprising:
automatically generating, with the computer processor, a written justification for the recommendation to accept or reject the potential comparable, wherein the written justification is specific to the potential comparable that has been analyzed.

3. The computer-implemented method of claim 2 , wherein the written justification is acceptable for justifying a transfer pricing use case.

4. The computer-implemented method of claim 1 , wherein the electronic data sources further include text obtained from a website of the comparable.

5. The computer-implemented method of claim 1 , further comprising:
generating, with the computer processor, probabilistic training data for the product classifier and the service classifier; and
training, with the computer processor, the product classifier and the service classifier using the probabilistic training data.

6. The computer-implemented method of claim 5 , wherein the probabilistic training data is generated using heuristic rules to generate heuristic labels.

7. The computer-implemented method of claim 6 , wherein the probabilistic training data is generated using an unsupervised label model.

8. The computer-implemented method of claim 1 , further comprising determining a risk of the potential comparable using a corresponding risk classifier.

9. A computer-implemented system for identifying comparables, the system comprising:
an electronic storage device; and
a machine learning algorithm operating on a computer processor that is programmed to:
collect, through an electronic interface, input data from a plurality of electronic data sources of a potential comparable for inclusion into a transfer pricing benchmarking set, wherein the transfer pricing benchmarking set is generated by a transfer pricing benchmarking activity which comprises search and selection of unrelated parties to establish a benchmark for the pricing of a cross-border transaction between one or more related parties under common ownership or control, wherein the electronic data sources include: a business description from a commercially available database, financial data of the potential comparable, and a SIC or NACE code associated with the potential comparable;
determine a function, product, and service of the potential comparable using a corresponding function classifier; product classifier, and service classifier;
receive, through the electronic interface, attributes of a tested party;
automatically execute a scoring process to calculate a similarity score for the potential comparable, wherein the similarity score represents a similarity between the potential comparable and the tested party;
automatically generate a recommendation to accept the potential comparable as an acceptable comparable for transfer pricing benchmarking, reject the potential comparable as a rejected comparable, or subject the potential comparable to further review;
generate, via the machine learning algorithm, synthetic training data for the machine learning algorithm by (1) predicting a plurality of labels and keyword counts from the plurality of electronic data sources, (2) generating heuristic labels for the predicted plurality of labels based on a plurality of heuristic rules, the rules based on experience with prior similar problems, and (3) generating probabilistic training labels for a semi-supervised deep learning model; and
improve the machine learning algorithm by using the synthetic training data and feedback from the automatically generated recommendation to train the machine learning algorithm.

10. The computer-implemented system of claim 9 , wherein the computer processor is further programmed to:
automatically generate a written justification for the recommendation to accept or reject the potential comparable, wherein the written justification is specific to the potential comparable that has been analyzed.

11. The computer-implemented system of claim 10 , wherein the written justification is acceptable for justifying a transfer pricing use case.

12. The computer-implemented system of claim 9 , wherein the electronic data sources further include text obtained from a website of the comparable.

13. The computer-implemented system of claim 9 , wherein the computer processor is further programmed to
generate probabilistic training data for the product classifier and the service classifier; and
train the product classifier and the service classifier using the probabilistic training data.

14. The computer-implemented system of claim 13 , wherein the probabilistic training data is generated using heuristic rules to generate heuristic labels.

15. The computer-implemented system of claim 14 , wherein the probabilistic training data is generated using an unsupervised label model.

16. The computer-implemented system of claim 9 , wherein the computer processor is further programmed to determine a risk of the potential comparable using a corresponding risk classifier.

17. A computer-implemented method for identifying comparables, the method comprising:
implementing a machine learning algorithm operating on a computer processor configured to:
(a) collect, through an electronic interface, input d

Assignees

• Kpmg Llp

Inventors

• Galginaitis Steven
• Braun Michel
• Cretcher Matthew
• Shea Brian
• Miller Mitchell
• Reitz Rebecca
• Schleicher Jillian
• Lee John Hyung
• Allen Marc Daniel

Classifications

• G06Q10/0635 Primary

  Risk analysis of enterprise or organisation activities · CPC title

• G06F16/9536

  Search customisation based on social or collaborative filtering · CPC title

• G06N5/01

  Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound · CPC title

• G06N5/04

  Inference or reasoning models · CPC title

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title