Personalized search library based on continual concept correlation
First Claim
1. A computing device to provide a personalized search library based on continual concept correlation, the computing device comprising:
- natural language analyzer circuitry to (i) receive event data representing content accessed by a user of a client computing device, wherein an order of the event data represents an order that the user of the client computing device accessed the content and (ii) analyze the event data to extract concepts of the content; and
correlation circuitry to (i) correlate the extracted concepts based on the order of the event data to generate a plurality of correlations between the extracted concepts, wherein each correlation is indicative of a relationship between a first extracted concept and a second extracted concept based on the order of the event data, (ii) adjust a weight associated with each extracted concept based on a frequency of the extracted concept occurring in the content to generate an adjusted weight, and (iii) store the correlations, adjusted weights, and extracted concepts in a concept model that identifies the relative correlation between each extracted concept and the adjusted weight associated with each extracted concept.
1 Assignment
0 Petitions
Accused Products
Abstract
A system, devices, and methods for providing a personalized search library based on continual concept correlation include a client computing device and a personalized content server. Content events representing content accessed or manipulated by a user of the client computing device are continually generated. Content associated with the content events is continually parsed and analyzed to extract main concepts. The extracted concepts are correlated and weighted into a concept model, based on the order of the content events. The concept model parallels the structure of the user'"'"'s memory. Data sources are continually searched for content relevant to a current context of the concept model. Relevant content is indexed according to the concept model. The relevant content may be made available to the user upon request or proactively. Relevant content may be cached for future use by the user. Other embodiments are described and claimed.
37 Citations
20 Claims
-
1. A computing device to provide a personalized search library based on continual concept correlation, the computing device comprising:
-
natural language analyzer circuitry to (i) receive event data representing content accessed by a user of a client computing device, wherein an order of the event data represents an order that the user of the client computing device accessed the content and (ii) analyze the event data to extract concepts of the content; and correlation circuitry to (i) correlate the extracted concepts based on the order of the event data to generate a plurality of correlations between the extracted concepts, wherein each correlation is indicative of a relationship between a first extracted concept and a second extracted concept based on the order of the event data, (ii) adjust a weight associated with each extracted concept based on a frequency of the extracted concept occurring in the content to generate an adjusted weight, and (iii) store the correlations, adjusted weights, and extracted concepts in a concept model that identifies the relative correlation between each extracted concept and the adjusted weight associated with each extracted concept. - View Dependent Claims (2, 3, 4, 5, 6, 7)
-
-
8. One or more non-transitory, machine readable media comprising a plurality of instructions that in response to being executed result in a computing device:
-
receiving event data representing content accessed by a user of a client computing device, wherein an order of the event data represents an order that the user of the client computing device accessed the content; analyzing the event data to extract concepts of the content; correlating the extracted concepts based on the order of the event data to generate a plurality of correlations between the extracted concepts, wherein each correlation is indicative of a relationship between a first extracted concept and a second extracted concept based on the order of the event data; adjusting a weight associated with each extracted concept based on a frequency of the extracted concept occurring in the content to generate an adjusted weight; and storing the correlations, adjusted weights, and extracted concepts in a concept model that identifies the relative correlation between each extracted concept and the adjusted weight associated with each extracted concept. - View Dependent Claims (9, 10, 11, 12, 13, 14)
-
-
15. A method to provide a personalized search library based on continual concept correlation, the method comprising:
-
receiving, by a computing device, event data representing content accessed by a user of a client computing device, wherein an order of the event data represents an order that the user of the client computing device accessed the content; analyzing, by the computing device, the event data to extract concepts of the content; correlating, by the computing device, the extracted concepts based on the order of the event data to generate a plurality of correlations between the extracted concepts, wherein each correlation is indicative of a relationship between a first extracted concept and a second extracted concept based on the order of the event data; adjusting, by the computing device, a weight associated with each extracted concept based on a frequency of the extracted concept occurring in the content to generate an adjusted weight; and storing, by the computing device, the correlations, adjusted weights, and extracted concepts in a concept model that identifies the relative correlation between each extracted concept and the adjusted weight associated with each extracted concept. - View Dependent Claims (16, 17, 18, 19, 20)
-
Specification