OPTIMIZING RESOURCE ALLOCATION TO A BID REQUEST RESPONSE BASED ON COGNITIVE ANALYSIS OF NATURAL LANGUAGE DOCUMENTATION

• US 20190317991A1
• Filed: 04/17/2018
• Published: 10/17/2019
• Est. Priority Date: 04/17/2018
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Abrahams, Faried, Hodge, Rashida A., Sivakumar, Gandhi, Dhondse, Amol A., Pikle, Anand
• Granted Patent: US 10,606,951 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 16/3346 using probabilistic model
• CPC Class Codes: G06F 16/9024 Graphs; Linked lists G06F16/9027 takes precedence
• CPC Class Codes: G06F 16/90332 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 16/9038 Presentation of query results
• CPC Class Codes: G06F 40/284 Lexical analysis, e.g. tokenisation or collocates
• CPC Class Codes: G06F 40/295 Named entity recognition
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06Q 10/101 Collaborative creation, e.g. joint development of products or services
• CPC Class Codes: G06Q 30/06 Buying, selling or leasing transactions

Abstract

Approaches presented herein enable optimizing resource allocation to a bid request response based on a cognitive analysis of natural language artifacts. More specifically, a bid request and a plurality of supporting artifacts in a natural language are obtained. A cognitive analysis of the request and supporting artifacts is performed to extract a set of information entities. The extracted information entities are normalized using a lexical-relations based graph database to classify the set of extracted information entities as standardized concepts. A subset of the set of the standardized concepts is identified as a set of parameters corresponding to a set of predetermined variables. Each variable of the set of predetermined variables is weighted according to a likelihood that the variable indicates a relevance of a resource. A probability that a particular resource is relevant is determined based on the weighting and that resource is assigned to the bid request response.

Claims

1. A method for optimizing resource allocation to a bid request response based on a cognitive analysis of natural language artifacts, comprising:

obtaining a request and a plurality of supporting artifacts in a natural language;

performing a cognitive analysis of the request and supporting artifacts to extract a set of information entities;

normalizing the extracted information entities using a lexical-relations based graph database to classify the set of extracted information entities as standardized concepts with which the set of extracted information entities are most closely associated;

identifying, for a portion of the request, at least a subset of the set of the standardized concepts, with which the set of extracted information entities are most closely associated, as a set of parameters corresponding with a set of predetermined variables;

weighting each variable of the set of predetermined variables according to a likelihood that the variable indicates a relevance of a resource to the portion of the request; and

assigning a particular resource to the bid request response in response to a probability that the particular resource is relevant to the portion of the request based on the weighted variables.

2. The method of claim 1, the method further comprising generating a list of probabilities that a set of resources are relevant to the portion of the request.

3. The method of claim 1, the determining a probability further comprising applying a first-order supervised multi-label learning classification algorithm to the weighted set of predetermined variables.

4. The method of claim 1, wherein the probability that the particular resource is relevant to the portion of the request has a variance and the method further comprises iterating the identifying, weighting, and determining until the variance is below a pre-defined threshold.

5. The method of claim 1, wherein the plurality of supporting artifacts comprises an artifact selected from the group consisting of:

historical data, public information, and correspondence.

6. The method of claim 1, wherein at least one variable of the set of predetermined variables is a variable selected from the group consisting of:

request type, deal size, techno-functional area of a deal, deal type, deal coverage, industry, sub-domain of an industry, techno-functional area of an industry, geography, client size, historical spending, inclination to a bidder, inclination to partners, bidder strength in a domain, sales organization size of a bidder, section of the request, and techno-functional entities of the request.

7. The method of claim 1, wherein the likelihood that the variable indicates a relevance of a resource to the portion of the request is based on historical data processed by a self-learning system.

8. A computer system for optimizing resource allocation to a bid request response based on a cognitive analysis of natural language artifacts, the computer system comprising:

a memory medium comprising program instructions;

a bus coupled to the memory medium; and

a processor, for executing the program instructions, coupled to a bid request analysis engine via the bus that when executing the program instructions causes the system to;

obtain a request and a plurality of supporting artifacts in a natural language;

perform a cognitive analysis of the request and supporting artifacts to extract a set of information entities;

normalize the extracted information entities using a lexical-relations based graph database to classify the set of extracted information entities as standardized concepts with which the set of extracted information entities are most closely associated;

identify, for a portion of the request, at least a subset of the set of the standardized concepts, with which the set of extracted information entities are most closely associated, as a set of parameters corresponding with a set of predetermined variables;

weight each variable of the set of predetermined variables according to a likelihood that the variable indicates a relevance of a resource to the portion of the request; and

assign a particular resource to the bid request response in response to a probability that the particular resource is relevant to the portion of the request based on the weighted variables.

9. The computer system of claim 8, the instructions further causing the system to generate a list of probabilities that a set of resources are relevant to the portion of the request.

10. The computer system of claim 8, the instructions further causing the system to apply a first-order supervised multi-label learning classification algorithm to the weighted set of predetermined variables.

11. The computer system of claim 8, wherein the probability that the particular resource is relevant to the portion of the request has a variance and the instructions further cause the system to iterate the identifying, weighting, and determining until the variance is below a pre-defined threshold.

12. The computer system of claim 8, wherein the plurality of supporting artifacts comprises an artifact selected from the group consisting of:

historical data, public information, and correspondence.

13. The computer system of claim 8, wherein at least one variable of the set of predetermined variables is a variable selected from the group consisting of:

request type, deal size, techno-functional area of a deal, deal type, deal coverage, industry, sub-domain of an industry, techno-functional area of an industry, geography, client size, historical spending, inclination to a bidder, inclination to partners, bidder strength in a domain, sales organization size of a bidder, section of the request, and techno-functional entities of the request.

14. The computer system of claim 8, wherein the likelihood that the variable indicates a relevance of a resource to the portion of the request is based on historical data processed by a self-learning system.

15. A computer program product for optimizing resource allocation to a bid request response based on a cognitive analysis of natural language artifacts, the computer program product comprising a computer readable hardware storage device, and program instructions stored on the computer readable hardware storage device, to:

obtain a request and a plurality of supporting artifacts in a natural language;

perform a cognitive analysis of the request and supporting artifacts to extract a set of information entities;

normalize the extracted information entities using a lexical-relations based graph database to classify the set of extracted information entities as standardized concepts with which the set of extracted information entities are most closely associated;

identify, for a portion of the request, at least a subset of the set of the standardized concepts, with which the set of extracted information entities are most closely associated, as a set of parameters corresponding with a set of predetermined variables;

weight each variable of the set of predetermined variables according to a likelihood that the variable indicates a relevance of a resource to the portion of the request; and

assign a particular resource to the bid request response in response to a probability that the particular resource is relevant to the portion of the request based on the weighted variables.

16. The computer program product of claim 15, the computer readable storage device further comprising instructions to generate a list of probabilities that a set of resources are relevant to the portion of the request.

17. The computer program product of claim 15, the computer readable storage device further comprising instructions to apply a first-order supervised multi-label learning classification algorithm to the weighted set of predetermined variables.

18. The computer program product of claim 15, wherein the probability that the particular resource is relevant to the portion of the request has a variance and the computer readable storage device further comprising instructions to iterate the identifying, weighting, and determining until the variance is below a pre-defined threshold.

19. The computer program product of claim 15, wherein at least one variable of the set of predetermined variables is a variable selected from the group consisting of:

request type, deal size, techno-functional area of a deal, deal type, deal coverage, industry, sub-domain of an industry, techno-functional area of an industry, geography, client size, historical spending, inclination to a bidder, inclination to partners, bidder strength in a domain, sales organization size of a bidder, section of the request, and techno-functional entities of the request.

20. The computer program product of claim 15, wherein the likelihood that the variable indicates a relevance of a resource to the portion of the request is based on historical data processed by a self-learning system.