Hybrid artificial intelligence system for semi-automatic patent infringement analysis

1. A computer-implemented method for performing patent claims analysis by software component executing on one or more servers comprising:

using preconditioning rules to determine relevant portions of a reference document, the preconditioning rules used to: exclude figures and tables from the reference document, and through a user interface, enable a user to resolve meanings of any acronyms found in the reference document or enter annotations for the reference document to provide further definitions or to indicate text weighted higher by the user during searching;

submitting a query to find a set of patents that are similar to the reference document;

receiving a set of matching patents from the query and using claim construction rules to determine how a meaning of claim elements of the patents will be ascertained;

using patent infringement rules and an artificial intelligence (AI) system to search the relevant portions of the reference document with the claim constructed claim elements of the patents returned from the search, wherein the claim construction rules and the patent infringement rules are based on legal statues of one or more countries; and

ranking the set of patents based on assigned confidence scores indicating a degree to which respective claim elements match the relevant portions of the reference document, and presenting the results to a client device.

2. The method of claim 1 , wherein searching the relevant portions of the reference document with the claim constructed claim elements of the patents further comprises:

inputting text of the reference document into the AI system to determine the meanings of words in the reference document;

inputting text of at least a portion of the patents returned from the patent search to the AI system to determine meanings of the claim elements based on the meanings of words in the corresponding patent; and

for each of the patents, performing, by the AI system, an element-by-element search to find matches between the claim elements and the relevant portions of the reference document based on the meanings of words.

3. The method of claim 2 , wherein performing the element-by-element search further comprises using, by the AI system, a machine learning (ML model), a natural language understanding (NLU) system, or both, to perform the element-by-element search.

4. The method of claim 3 , further comprising inputting the text of the reference document and the text of the patents into the ML model and the NLU system in parallel.

5. The method of claim 3 , further comprising inputting the text of the reference document and the text of the patents into the ML model and the NLU system serially where the output of one is used as input to the other.

6. The method of claim 2 , wherein inputting the text of at least a portion of the patents returned from the patent search to the AI system further comprises parsing each of the patents to extract a specification and to extract the claims therein, and passing the specification and the claims to the AI system with a corresponding patent identifier.

7. The method of claim 1 , further comprising basing the claim construction rules and the patent infringement rules on statutes, codes, or case law that are selectable by a user.

8. A system, comprising:

a hybrid artificial intelligence (AI) system in communication with a patent search provider and a client device over a network, the hybrid AI system to:

receive a search request from the client device to find patents that read on a text description provided in a reference document;

determine relevant portions of the reference document to narrow a focus of subsequent analysis by using preconditioning rules that: exclude figures and tables from the reference document, and through a user interface, enable a user to resolve meanings of any acronyms found in the reference document or enter annotations for the reference document to provide further definitions or to indicate text weighted higher by the user during searching;

perform a patent search on the reference document to find a set of patents that are similar to the reference document

and an AI system in communication with the hybrid AI system, the AI system to

input text of the reference document to determine the meanings of words in the reference document;

input text of at least a portion of the patents returned from the patent search to determine meanings of the claim elements based on the meanings of words in corresponding patent;

for each of the patents, perform an element-by-element search to find matches between the claim elements and the relevant portions of the reference document based on the meanings of words; and

output to the client device a list of the patents having claims elements that match the text of the relevant portions of the reference document along with element confidence scores or claim confidence scores.

9. The system of claim 8 , wherein the AI system uses a machine learning (ML model), a natural language understanding (NLU) system, or both to perform the element-by-element search.

10. The system of claim 9 , wherein the text of the reference document and the text of the patents are input into the ML model and the NLU system in parallel.

11. The system of claim 9 , wherein the text of the reference document and the text of the patents are input into the ML model and the NLU system serially where the output of one is used as input to the other.

12. The system of claim 8 , wherein the patents returned from the patent search are parsed to extract a specification and to extract the claims before being input to the AI system.

13. A non-transitory computer-readable medium containing program instructions for performing patent claims analysis, the program instructions for:

using preconditioning rules to determine relevant portions of a reference document, the preconditioning rules used to: exclude figures and tables from the reference document, and through a user interface, enable a user to resolve meanings of any acronyms found in the reference document or enter annotations for the reference document to provide further definitions or to indicate text weighted higher by the user during searching;

submitting a query to find a set of patents that are similar to the reference document;

receiving a set of matching patents from the query and using claim construction rules to determine how a meaning of claim elements of the patents will be ascertained;

using patent infringement rules and an artificial intelligence (AI) system to search the relevant portions of the reference document with the claim constructed claim elements of the patents returned from the search, wherein the claim construction rules and the patent infringement rules are based on legal statues of one or more countries; and

ranking the set of patents based on assigned confidence scores indicating a degree to which respective claim elements match the relevant portions of the reference document, and presenting the results to a client device.

14. The non-transitory computer-readable medium of claim 13 , wherein searching the relevant portions of the reference document with the claim constructed claim elements of the patents further comprises:

inputting text of the reference document into the AI system to determine the meanings of words in the reference document;

inputting text of at least a portion of the patents returned from the patent search to the AI system to determine meanings of the claim elements based on the meanings of words in the corresponding patent; and

for each of the patents, performing, by the AI system, an element-by-element search to find matches between the claim elements and the relevant portions of the reference document based on the meanings of words.

15. The non-transitory computer-readable medium of claim 14 , wherein performing the element-by-element search further comprises using, by the AI system, a machine learning (ML model), a natural language understanding (NLU) system, or both, to perform the element-by-element search.

16. The non-transitory computer-readable medium of claim 15 , further comprising inputting the text of the reference document and the text of the patents into the ML model and the NLU system in parallel.

17. The non-transitory computer-readable medium of claim 15 , further comprising inputting the text of the reference document and the text of the patents into the ML model and the NLU system serially where the output of one is used as input to the other.

18. The non-transitory computer-readable medium of claim 14 , wherein inputting the text of at least a portion of the patents returned from the patent search to the AI system further comprises parsing each of the patents to extract a specification and to extract the claims therein, and passing the specification and the claims to the AI system with a corresponding patent identifier.