1. A method for determining whether a document would be obvious to a person having ordinary skill in the art: Receiving a first document whose obviousness is to be determined; Receiving two or more second documents on basis of which obviousness is to be determined; Analyzing the first document to develop a first understanding of the first document; Generating a plurality of features from the first understanding; Analyzing the two or more second documents to develop a plurality of second understandings; Generating a plurality of features from each of the plurality of second understandings; Individually comparing the features of each of the plurality of second understandings with the features of first document; Determining a subset from the second or more second documents, wherein the features generated from the subset includes all the features generated from the first document; Determining a probability of first document being found obvious to a person having ordinary skill in the art, wherein the probability is calculated based on number of documents in the subset, and wherein the determination is made by an artificial model trained to make such determination.

2. A system for determining whether a document would be obvious to a person having ordinary skill in the art: A document receiving module for receiving as input a first document, whose obviousness is to be determined as well as a second set of documents, on whose basis the obviousness is to be determined; A feature extraction AI model for generating features from first document as well as second set of documents; A subset generation AI model for comparing features generated from first document with features generated from features generated from second set of documents; A determination AI model used for determining the probability that the first document would be obvious to a person having ordinary skill in the art; An output module for outputting the probability.

3. The system of claim 2, wherein the subset generating module for generating a subset, wherein the features generated from the subset includes all the features generated from the first document.

4. The system of claim 2 further including a feedback receiving module, wherein feedback received by the feedback receiving module is used to enhance one or more of first AI module, second AI module and obviousness determining AI module.

5. The system of claim 2, wherein the first AI model, the second AI model, and the obviousness determining AI model are all trained before.

6. The system of claim 2, wherein the obviousness determining AI model is pre-trained using a database containing one or more of legal documents, human expert opinion and/or USPTO judgements.

Description:TECHNICAL FIELD

[0001] The present invention is directed towards a method and a system for determining whether a potential invention may be obvious to a person having ordinary skill the art in view of current state of the art.

5

DESCRIPTION OF THE RELATED ART

[0002] Intellectual property rights are the most effective tools for promoting intellectual endeavor among masses. One of the oldest forms of Intellectual Property rights is Patent. Patent is an exclusionary right that has existed since at least medieval

10 period. And by some accounts, the oldest Patent system might even be older, going as far back as the Roman empire.

[0003] In the current scenario, two universally accepted conditions for a patent to be granted are novelty and non-obviousness.

[0004] Concept of novelty is very well defined. An invention is considered novel

15 when there does not exist any single public disclosure in which the complete invention is described.

[0005] Concept of non-obviousness or, rather, obviousness is a little more complicated. Broadly, an invention is considered obvious, if a Person-Having-Ordinary-Skill-In-
The-Art (PHOSITA) would be able to arrive at the invention based solely on publicly

20 available disclosures, without application of any specific skill.

[0006] PHOSITA is a hypothetical person, generally considered as possessing common sense. These qualities of PHOSITA make determination of whether any potentially new invention is obvious or not a subjective determination. The subjectiveness of determination can cause various issues at various stages.

25 [0007] An example of one of issues, at pre-filing stage, is determining whether an invention should be filed at PTO or not. Filing and prosecuting a patent application can

2

be a very expensive task, more so if the applications are to be filed in many different jurisdictions. To save these expenses, the inventors generally invest in prior art searches, and only when the invention is considered novel and non-obvious, do the inventors file the applications with respective patent offices.

5 [0008] The prior art searches are usually performed by experts and therefore, the opinion on obviousness of an invention may vary from individual to individual. At this stage, the subjectiveness may cause inventor to invest in filing of an application, whose probability of being found obvious by the PTO is very high.

[0009] Another stage where subjectiveness of obviousness is presents an issue is the

10 Examiner appointed by the PTO. At a PTO, multiple examiners are available for prosecution of similar applications. Since opinion on obviousness of an invention may vary from examiner to examiner, which examiner gets assigned which application may ultimately end up deciding which application gets granted or rejected.

[0010] In view of the foregoing, a method and corresponding system is disclosed to

15 address the abovementioned issues in the art.

SUMMARY

[0011] In view of the foregoing, an embodiment herein provides a method for determining probability of a document being found obvious in view of state of art. The

20 method comprises steps of receiving a document, the invention document, whose obviousness is to be determined from a user, receiving a set of documents representing state of art, prior art documents, generating a first set of features from invention document, generating a second set of features from prior art documents, wherein the second set contains plurality of subsets of features, wherein each subset of features

25 corresponds to one of the prior art documents. The method further comprises determining a subset of prior art documents, wherein the set which is union of subsets of features corresponding to the subset of prior art documents include all the features

3

of the invention document. The method further comprises determining probability of invention document being found obvious.

[0012] In one embodiment, the determination of probability is made using an Artificial Intelligence module, which has been trained on one or more sets of data. One of the

5 sets of data could be applications which were not granted based on obviousness, and applications which were able to overcome obviousness-based rejections. Another set of data could be court or appellate decisions. Another set of data could be the opinion of multiple experts.

[0013] In this document, we interchangeably use the terms “Artificial Intelligence”,

10 “AI”, “Machine learning”, “ML”, “Neural Network”, “NN”. All these terms refer to software capable of mimicking human intelligence and human cognitive functions.

[0014] In one embodiment, the step of generating set(s) of features is performed an Artificial Intelligence module.

[0015] In one embodiment, the step of determining subsets is performed by an Artificial

15 Intelligence module.

BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The embodiments herein will be better understood from the following detailed description with reference to the drawings, in which:

[0017] FIG. 1 illustrates a first exemplary system of the invention.

20

25

4

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

[0018] The embodiments herein and the various features and advantageous details thereof are explained more fully with reference to the non-limiting embodiments that are illustrated in the accompanying drawings and detailed in the following description.

5 Descriptions of well-known components and processing techniques are omitted so as not to unnecessarily obscure the embodiments herein. The examples used herein are intended merely to facilitate an understanding of ways in which the embodiments herein may be practiced and to further enable those of skill in the art to practice the

embodiments herein. Accordingly, the examples should not be construed as limiting

10 the scope of the embodiments herein.

[0019] As mentioned, there remains a need for a system and a method determining probability of a document being found obvious. Referring now to FIG. 1, where similar reference characters denote corresponding features consistently throughout the figures, there are shown preferred embodiments.

15 [0020] FIG. 1 illustrates a first exemplary implementation of a system capable of determining probability of a document being found obvious according to an embodiment herein.

[0021] System 100 includes Input Module 110, Feature Extraction Module 120, Subset

Generation Module 130, Determination Module 140, and Output Module 150. The

20 system may be implemented either on a standalone computing machine or provided as service across a network. Even though the type of network is not essential to the invention, some non-limiting examples of network are LAN, WAN, Internet, Intranet, Cellular Network, Decentralized Network etc.

[0022] The Input Module 110 includes a Document Input Submodule 111 and a

25 Feedback Input Submodule 112. The Document Input Submodule 111 allows a user to input invention documents as well as prior art documents to the system, and the Feedback Input Submodule 112 allows a user to provide feedback in response to any

5

information provided to user by the Output Module 150. Here input encompasses all possible definitions of the word. As one non-limiting example, the Document Input Submodule 111 allows the user to type and/or dictate the whole invention. As another non-limiting example, the Document Input Submodule 111 allows users to upload

5 digital files of invention document as well prior art documents. As another non-limiting example, the Document Input Submodule 111 allows user to provide Universal Resource Locator (URL) address of invention document as well as prior art documents. When the user inputs invention document as well as prior art documents by uploading
digital files, the format of files may be, as non-limiting example, one of MS-Word,

10 MS-Excel, PDF, JSON etc. In one embodiment, the Document Input Submodule 111 allows users to navigate the location of invention documents as well as prior art documents. In one embodiment, the method of inputting invention document and prior art documents are different. In another embodiment, the method of inputting different prior art documents is different. In another embodiment invention document as well as

15 prior art documents are input in same manner. In one embodiment the Input Module 110 validates the digital files. The Input Module 110 provides the invention document as well the prior art documents to the Feature Extraction Module 120.

[0023] The Feature Extraction Module 120 is an AI module that mimics the function

of understanding a document of human mind. The Feature Extraction Module 120

20 processes the invention document as well as all the prior art documents to generate one set of features for the invention document as well as one set of features for each of the prior art documents. In one embodiment, after generating features of invention

document as well as every prior art document, the Feature Extraction Module 120 presents the generated features to user, using the output module 150. The user may

25 provide feedback on the generated features using the feedback input submodule 112. In case the user feedback is negative, a new set of features are generated for the corresponding document. In case the feedback is positive, the features are provided to

6

the Subset Generation Module 130. Irrespective of feedback being positive or negative, the feedback is also used for enhancing the Feature Extraction Module 120 in real-time.

[0024] In one embodiment, the Feature Extraction Module 120 is trained prior to becoming part of the system 100.

5 [0025] The Subset Generation Module 130 takes all sets of features corresponding to the prior art documents. The Subset Generation Module 130 then outputs a subset of documents. The generated subset of documents is provided to the Determination Module 140. To generate the subset, the Subset Generation Module 130 creates all

possible combinations of prior art documents. Next a union set of sets of features of

10 prior art documents in each combination is generated. This union set is compared to the set of features of the invention document. In one embodiment, all the combinations of prior art documents whose union set contains all the features of the invention document present is output to the Determination Module 140. In another embodiment, only a certain number of such combinations are output to the Determination Module

15 140. In another embodiment, the combination of prior arts with the least number of prior arts is selected for output. In case there are plurality of such combinations, either all combinations are evaluated, or the user is presented with the combinations using the Output Module 150 and asked to select combinations using input module 110. The user selection may also be used to train the Subset Generation Module 130.

20 [0026] The Determination Module 140 determines the probability of the invention document being found obvious. The Determination Module 140 is previously trained using one or more of the following datasets: A dataset containing court rulings where

the court ruled on obviousness on an invention document in view of state of art, A dataset containing expert opinions, and A dataset containing Examiner rejections on

25 obviousness grounds.

[0027] The cases where applicant overcame the obviousness rejections using claim amendments or the applicant abandoned the application after obviousness rejections

7

would be considered as a case where the Examiner correctly determined the invention to be obvious. In cases where applicant overcame the obviousness rejections by explaining why the invention was not obvious in view of asserted documents would be considered as a case where the Examiner incorrectly determined the invention to be
5 obvious.

[0028] In case where only 1 subset is evaluated, the Determination Module 140 informs the user of probability of invention being determined obvious using the Output Module 150. In case where plurality of subsets is evaluated, the Determination Module 140
may inform, using the Output Module 150, the user of subset with highest probability

10 of invention being determined obvious. In case where plurality of subsets is evaluated, the Determination Module 140 may inform the user, using the Output Module 150 of probability of invention being determined obvious for every subset.

[0029] Herein an embodiment is envisioned in which the Input Module 110 does not include the Feedback Input Submodule 112. In this embodiment, none of the modules

15 asks the user to provide any feedback to the system.

[0030] FIG. 2 illustrates a method of invention.

[0031] In step 201, the Input Module 110 presents a User Interface to a user that allows the user to input Invention document as well as prior art documents. For the invention
it is not important whether prior art documents are input first, or invention document

20 is input first, or all the documents are input together, or which method is implemented for inputting the documents. All the prior art documents as well as the invention document are provided to the Feature Extraction Module 120.

[0031] the Input module 110 provides the documents to the Feature Extraction Module 120.

25 [0032] In step 202, the Feature Extraction Module 120 extracts the features from Input document as well as all the prior art documents.

8

[0033] the Feature Extraction Module 120 outputs the features to the Subset Generation Module 130.

[0034] In one embodiment, a step of presenting the features to user and collecting feedback may be implemented between step 202 and step 203.

5 [0035] In step 203, the Subset Generation Module 130 generates one or more subset of prior art documents such that all the features of invention documents are presents in the union set of sets of features of the prior art document in the subset.

[0036] The Subset Generation Module 130 provides one or more subsets to the Determination Module 140.

10 [0037] In one embodiment, a step of presenting the subsets to user and collecting feedback maybe implemented between step 203 and step 204.

[0038] In step 204, the Determination Module 140 calculates the probability of invention being obvious in view of the prior art documents. The calculated probability is presented to the user using the Output Module 150.

15 [0039] The foregoing description of the specific embodiments has fully revealed the general nature of the embodiments herein that others can, by applying current knowledge, readily modify and/or adapt for various applications such specific embodiments without departing from the generic concept, and, therefore, such

adaptations and modifications should and are intended to be comprehended within the

20 meaning and range of equivalents of the disclosed embodiments. It is to be understood that the phraseology or terminology employed herein is for the purpose of description and not of limitation. Therefore, while the embodiments herein have been described in terms of preferred embodiments, those skilled in the art will recognize that the embodiments herein can be practiced with modification within the spirit and scope. , Claims:I/We Claim:
1. A method for determining whether a document would be obvious to a person having ordinary skill in the art:
Receiving a first document whose obviousness is to be determined;
Receiving two or more second documents on basis of which obviousness is to be determined;
Analyzing the first document to develop a first understanding of the first document;
Generating a plurality of features from the first understanding;
Analyzing the two or more second documents to develop a plurality of second understandings;
Generating a plurality of features from each of the plurality of second understandings;
Individually comparing the features of each of the plurality of second understandings with the features of first document;
Determining a subset from the second or more second documents, wherein the features generated from the subset includes all the features generated from the first document;
Determining a probability of first document being found obvious to a person having ordinary skill in the art, wherein the probability is calculated based on number of documents in the subset, and wherein the determination is made by an artificial model trained to make such determination.
2. A system for determining whether a document would be obvious to a person having ordinary skill in the art:
A document receiving module for receiving as input a first document, whose obviousness is to be determined as well as a second set of documents, on whose basis the obviousness is to be determined;
A feature extraction AI model for generating features from first document as well as second set of documents;
A subset generation AI model for comparing features generated from first document with features generated from features generated from second set of documents;
A determination AI model used for determining the probability that the first document would be obvious to a person having ordinary skill in the art;
An output module for outputting the probability.
3. The system of claim 2, wherein the subset generating module for generating a subset, wherein the features generated from the subset includes all the features generated from the first document.
4. The system of claim 2 further including a feedback receiving module, wherein feedback received by the feedback receiving module is used to enhance one or more of first AI module, second AI module and obviousness determining AI module.
5. The system of claim 2, wherein the first AI model, the second AI model, and the obviousness determining AI model are all trained before.
6. The system of claim 2, wherein the obviousness determining AI model is pre-trained using a database containing one or more of legal documents, human expert opinion and/or USPTO judgements.