Predictive intelligent softbots on the cloud
A system and method for generating a prediction are disclosed. In one embodiment, the method includes receiving a plurality of time-based documents; receiving a user query including a time period of interest defining a subset of the time-based documents from which to generate a prediction; and a plurality of cloud-based software agents classifying the subset of the plurality of time-based documents into a plurality of classes for the plurality entities, wherein the plurality of cloud-based software agents intercommunicate using distributed processing, wherein each of the plurality of cloud-based software agents is dedicated to one of the entities in the plurality of entities; and generating, using at least one machine learning method, the prediction based on the subset of the plurality of time-based documents for at least one of a plurality of categories. However, other embodiments are disclosed.
1. A method comprising:
receiving, using one or more processors, a plurality of time-based documents;
receiving, by a Software as a Service cloud-based server, a user query including a time period of interest defining a subset of the plurality of time-based documents from which to generate a prediction;
determining to activate a plurality of cloud-based software agents, which execute on a plurality of remote computational devices geographically distributed across world geography, based on a plurality of entities associated with the user query, wherein each of the plurality of cloud-based software agents is dedicated to a single entity within the plurality of entities associated with the user query;
classifying, by executing the plurality of cloud-based software agents on the plurality of remote computational devices, the subset of the plurality of time-based documents into a plurality of classes for the plurality of entities, wherein the plurality of cloud-based software agents intercommunicate using distributed processing; and
generating, by executing the plurality of cloud-based software agents on the plurality of remote computational devices, using at least one machine learning method, the prediction based on the subset of the plurality of time-based documents for at least one of a plurality of categories.
2. The method as claimed in claim 1 , wherein the plurality of cloud-based software agents include one or more of an intelligent software agent, an automated software agent, and an intelligent automated software agent.
3. The method as claimed in claim 1 , wherein a first cloud-based software agent of the plurality of cloud-based software agents continuously updates its corresponding plurality of machine learning parameters.
4. The method as claimed in claim 3 , wherein the plurality of machine learning parameters include functional parameters derived from a statistical learning on a plurality of features extracted from the subset of the plurality of time-based documents using the at least one machine learning method.
5. The method as claimed in claim 4 , wherein the functional parameters are derived using Support Vector Machine classification as the at least one machine learning method.
6. The method as claimed in claim 1 , wherein the plurality of time-based documents comprise one or more of news stories, news blogs, and web documents about the plurality of entities.
7. The method as claimed in claim 1 , wherein the plurality of time-based documents are received from one or more of a plurality of news vendors and by mining a relevant plurality of hyperlink texts available on a world wide web, and wherein the user query includes a time period and the subset of the plurality of time-based documents is determined as those time-based documents associated with a time between a current time and a past time, the past time being calculated by subtracting the time period from the current time.
8. The method as claimed in claim 1 , wherein the plurality of entities comprise at least one of: a business organization, an individual, a community, and a group.
9. The method as claimed in claim 1 , wherein the plurality of time-based documents are classified using at least one of a natural language processing method and a machine learning method.
10. The method as claimed in claim 1 , wherein the plurality of categories include at least one of: a leadership change, a recent acquisition, a recent merger, and an intellectual property portfolio.
11. The method as claimed in claim 1 , wherein the plurality of cloud-based software agents communicate with one another using a message passing in a form of predicate logic.
12. The method of claim 1 , wherein the time period of interest establishes a time period from a current date to a past date, wherein a time-based document is associated with a time, the time associated with one or more of generation of the time-based document, transmission of the time-based document, receipt of the time-based document, publication of the time-based document, and an event described by the time-based document.
13. A method comprising:
receiving, using one or more processors, a time-based query from a user;
determining a classification criterion, a relationship criterion, and a time period of interest from the time-based query;
obtaining at least one second classification criterion, at least one second relationship criterion, a plurality of concepts, and a plurality of entities from a knowledge base;
classifying a plurality of documents based on the classification criterion, the at least one second classification criterion, and the time-based query using a distributed machine learning method;
extracting the plurality of entities for the relationship criterion, the at least one second relationship criterion from a plurality of classified documents based on the classification criterion, the at least one second classification criterion, and the time-based query using a distributed processing;
generating a plurality of data representations between the relationship criterion and said plurality of entities based on said classification criterion and the time-based query using the plurality of classified documents; and
applying a distributed processing method to manipulate said plurality of data representations for generating a visualization of a prediction responsive to the time-based query, wherein the prediction captures a combined influence existing between the classification criterion, the at least one second classification criterion, the relationship criterion, and the at least one second relationship criterion.
14. The method as claimed in claim 13 , wherein one or more of the classification criterion and the at least one second classification criterion include at least one of: a leadership change, a recent acquisition, a recent merger, and an intellectual property portfolio.
15. The method as claimed in claim 13 , wherein the time-based query is a query involving the plurality of entities mentioned in the plurality of documents during the time period of interest.
16. The method as claimed in claim 13 , wherein the classifying of the plurality of documents uses at least one machine learning method.
17. The method as claimed in claim 13 , wherein the relationship criterion and the at least one second relationship criterion include a relationship among the plurality of entities, wherein the relationship is an association between said entities and said criterion.
18. The method as claimed in claim 13 , wherein the extracting the plurality of entities utilizing the relationship criterion use at least one natural language processing method and said plurality of entities include one or more of an organization, an executive, an organization's product, an organization's revenue, an organization's share, and an executive's role.
19. The method as claimed in claim 13 , wherein the plurality of influential entities comprise at least one of: an organization, a group, a community, and an individual.
20. The method as claimed in claim 13 , wherein the plurality of concepts include real time events associated with said relationship criterions and said classification criterion.
21. The method as claimed in claim 13 , wherein the distributed processing comprises a map-reduce framework of Hadoop.
22. The method as claimed in claim 13 , wherein the visualization comprises one or more of a text and a graph.
23. The method as claimed in claim 13 , wherein the combined influence is an influence existing between said entities, said concepts, and said classification criterions.
24. A system comprising;
a processing device implementing a Software as a Service cloud-based server for receiving a user query including a time period of interest defining a subset of a plurality of received time-based documents from which to generate a prediction; and
a plurality of remote computational devices geographically distributed across world geography, the plurality of remote computational devices implementing a plurality of cloud-based software agents activated based on a plurality of entities associated with the user query, wherein each of the plurality of cloud-based software agents is dedicated to a single entity within the plurality of entities associated with the user query, the plurality of cloud-based software agents for classifying the subset of the plurality of time-based documents into a plurality of classes for the plurality of entities, wherein the plurality of cloud-based software agents intercommunicate using distributed processing, the plurality of cloud-based software agents generating, using at least one machine learning method, the prediction based on the subset of the plurality of time-based documents for at least one of a plurality of categories.
25. The system of claim 24 , wherein the plurality of cloud-based software agents include one or more of an intelligent software agent, an automated software agent, and an intelligent automated software agent.
26. The system of claim 24 , wherein a first cloud-based software agent of the plurality of cloud-based software agents continuously updates its corresponding plurality of machine learning parameters.
27. The system of claim 26 , wherein the plurality of machine learning parameters include functional parameters derived from a statistical learning on a plurality of features extracted from the subset of the plurality of time-based documents using the at least one machine learning method.
28. The system of claim 27 , wherein the functional parameters are derived using Support Vector Machine classification as the at least one machine learning method.
29. The system of claim 24 , wherein the plurality of time-based documents comprise one or more of news stories, news blogs, and web documents about the plurality of entities.
30. The system of claim 24 , wherein the plurality of time-based documents are received from one or more of a plurality of news vendors and by mining a relevant plurality of hyperlink texts available on a world wide web, and wherein the user query includes a time period and the subset of the plurality of time-based documents is determined as those time-based documents associated with a time between a current time and a past time, the past time being calculated by subtracting the time period from the current time.
31. The system of claim 24 , wherein the plurality of entities comprise at least one of: a business organization, an individual, a community, and a group.
32. The system of claim 24 , wherein the plurality of time-based documents are classified using at least one of a natural language processing method and a machine learning method.
33. The system of claim 24 , wherein the plurality of categories include at least one of: a leadership change, a recent acquisition, a recent merger, and an intellectual property portfolio.
34. The system of claim 24 , wherein the plurality of cloud-based software agents communicate with one another using a message passing in a form of predicate logic.
35. The system of claim 24 , wherein the time period of interest establishes a time period from a current date to a past date, wherein a time-based document is associated with a time, the time associated with one or more of generation of the time-based document, transmission of the time-based document, receipt of the time-based document, publication of the time-based document, and an event described by the time-based document.
36. A system comprising;
a memory; and
a processor operatively coupled to the memory, the processor configured to perform the steps of:
receiving a time-based query from a user;
determining a classification criterion, a relationship criterion, and a time period of interest from the time-based query;
obtaining at least one second classification criterion, at least one second relationship criterion, a plurality of concepts and a plurality of entities from a knowledge base;
classifying a plurality of documents based on the classification criterion, the at least one second classification criterion, and the time-based query using a distributed machine learning method;
extracting the plurality of entities for the relationship criterion, the at least one second relationship criterion from a plurality of classified documents based on the classification criterion, the at least one second classification criterion, and the time-based query using a distributed processing;
generating a plurality of data representations between the relationship criterion and said plurality of entities based on said classification criterion and the time-based query using the plurality of classified documents; and
applying a distributed processing method to manipulate said plurality of data representations for generating a visualization of a prediction responsive to the time-based query, wherein the prediction captures a combined influence existing between the classification criterion, the at least one second classification criterion, the relationship criterion, and the at least one second relationship criterion.
37. The system of claim 36 , wherein one or more of the classification criterion and the at least one second classification criterion include at least one of: a leadership change, a recent acquisition, a recent merger, and an intellectual property portfolio.
38. The system of claim 36 , wherein the time-based query is a query involving the plurality of entities mentioned in the plurality of documents during the time period of interest.
39. The system of claim 36 , wherein the classifying of the plurality of documents uses at least one machine learning method.
40. The system of claim 36 , wherein the relationship criterion and the at least one second relationship criterion include a relationship among the plurality of entities, wherein the relationship is an association between said entities and said criterion.
41. The system of claim 36 , wherein the extracting the plurality of entities utilizing the relationship criterion use at least one natural language processing method and said plurality of entities include one or more of an organization, an executive, an organization's product, an organization's revenue, an organization's share, and an executive's role.
42. The system of claim 36 , wherein the plurality of entities comprise at least one of: an organization, a group, a community, and an individual.
43. The system of claim 36 , wherein the plurality of concepts include real time events associated with said relationship criterions and said classification criterion.
44. The system of claim 36 , wherein the distributed processing comprises a map-reduce framework of Hadoop.
45. The system of claim 36 , wherein the visualization comprises one or more of a text and a graph.
46. The system of claim 36 , wherein the combined influence is an influence existing between said entities, said concepts, and said classification criterions.
47. A non-transitory computer-readable medium storing instructions that, when executed by one or more processors, performs steps comprising:
receiving a plurality of time-based documents;
receiving a user query including a time period of interest defining a subset of the plurality of time-based documents from which to generate a prediction;
determining to activate a plurality of cloud-based software agents, which execute on a plurality of remote computational devices geographically distributed across world geography, based on a plurality of entities associated with the user query, wherein each of the plurality of cloud-based software agents is dedicated to a single entity within the plurality of entities associated with the user query;
classifying, by executing the plurality of cloud-based software agents, the subset of the plurality of time-based documents into a plurality of classes for the plurality of entities, wherein the plurality of cloud-based software agents intercommunicate using distributed processing; and
generating, by executing the plurality of cloud-based software agents on the plurality of remote computational devices, using at least one machine learning method, the prediction based on the subset of the plurality of time-based documents for at least one of a plurality of categories.
48. The computer-readable medium of claim 47 , wherein the plurality of cloud-based software agents include one or more of an intelligent software agent, an automated software agent and an intelligent automated software agent.
49. The computer-readable medium of claim 47 , wherein a first cloud-based software agent of the plurality of cloud-based software agents continuously updates its corresponding plurality of machine learning parameters.
50. The computer-readable medium of claim 49 , wherein said plurality of machine learning parameters include functional parameters derived from a statistical learning on a plurality of features extracted from the subset of the plurality of time-based documents using the at least one machine learning method.
51. The computer-readable medium of claim 50 , wherein the functional parameters are derived using Support Vector Machine classification as the at least one machine learning method.
52. The computer-readable medium of claim 47 , wherein the plurality of time-based documents comprise one or more of news stories, news blogs, and web documents about the plurality of entities.
53. The computer-readable medium of claim 47 , wherein the plurality of time-based documents are received from one or more of a plurality of news vendors and by mining a relevant plurality of hyperlink texts available on a world wide web, and wherein the user query includes a time period and the subset of the plurality of time-based documents is determined as those time-based documents associated with a time between a current time and a past time, the past time being calculated by subtracting the time period from the current time.
54. The computer-readable medium of claim 47 , wherein the plurality of entities comprise at least one of: a business organization, an individual, a community, and a group.
55. The computer-readable medium of claim 47 , wherein the plurality of time-based documents are classified using at least one of a natural language processing method and a machine learning method.
56. The computer-readable medium of claim 47 , wherein the plurality of categories include at least one of: a leadership change, a recent acquisition, a recent merger, and an intellectual property portfolio.
57. The computer-readable medium of claim 47 , wherein the plurality of cloud-based software agents communicate with one another using a message passing in a form of predicate logic.
58. The computer-readable medium of claim 47 , wherein the time period of interest establishes a time period from a current date to a past date, wherein a time-based document is associated with a time, the time associated with one or more of generation of the time-based document, transmission of the time-based document, receipt of the time-based document, publication of the time-based document and an event described by the time-based document.
59. A non-transitory computer-readable medium storing instructions that, when executed by one or more processors, performs steps comprising:
receiving a time-based query from a user;
determining a classification criterion, a relationship criterion, and a time period of interest from the time-based query;
obtaining at least one second classification criterion, at least one second relationship criterion, a plurality of concepts, and a plurality of entities from a knowledge base;
classifying a plurality of documents based on the classification criterion, the at least one second classification criterion, and the time-based query using a distributed machine learning method;
extracting the plurality of entities for the relationship criterion, the at least one second relationship criterion from a plurality of classified documents based on the classification criterion, the at least one second classification criterion, and the time-based query using a distributed processing;
generating a plurality of data representations between the relationship criterion and said plurality of entities based on said classification criterion and the time-based query using the plurality of classified documents; and
applying a distributed processing method to manipulate said plurality of data representations for generating a visualization of a prediction responsive to the time-based query, wherein the prediction captures a combined influence existing between the classification criterion, the at least one second classification criterion, the relationship criterion, and the at least one second relationship criterion.
60. The computer-readable medium of claim 59 , wherein one or more of the classification criterion and the at least one second classification criterion include at least one of: a leadership change, a recent acquisition, a recent merger, and an intellectual property portfolio.
61. The computer-readable medium of claim 59 , wherein the time-based query is a query involving the plurality of entities mentioned in the plurality of documents during the time period of interest.
62. The computer-readable medium of claim 59 , wherein the classifying of the plurality of documents uses at least one machine learning method.
63. The computer-readable medium of claim 59 , wherein the relationship criterion and the at least one second relationship criterion include a relationship among the plurality of entities, wherein the relationship is an association between said entities and said criterion.
64. The computer-readable medium of claim 59 , wherein the extracting the plurality of entities utilizing the relationship criterion use at least one natural language processing method and said plurality of entities include one or more of an organization, an executive, an organization's product, an organization's revenue, an organization's share, and an executive's role.
65. The computer-readable medium of claim 59 , wherein the plurality of entities comprise at least one of: an organization, a group, a community, and an individual.
66. The computer-readable medium of claim 59 , wherein the plurality of concepts include real time events associated with said relationship criterions and said classification criterion.
67. The computer-readable medium of claim 59 , wherein the distributed processing comprises a map-reduce framework of Hadoop.
68. The computer-readable medium of claim 59 , wherein the visualization comprises one or more of a text and a graph.
69. The computer-readable medium of claim 59 , wherein the combined influence is an influence existing between said entities, said concepts, and said classification criterions.