Browse Prior Art Database

Predictive Caching System and Method Basing on User Behavior and Chracter Combination Set Trend

IP.com Disclosure Number: IPCOM000247152D
Publication Date: 2016-Aug-11
Document File: 9 page(s) / 155K

Publishing Venue

The IP.com Prior Art Database

Abstract

Our idea to provide a predictive caching system and method based on user count and behaviour weight in time interval. A Tag Group System is applied for grouping tags related to Content in hierarchy structure. The tag of content a user ever visits is weighted basing on this user’s behaviour weight. With analysing the user account and tag weight of a Tagged User Group which is composed of users who visit the content with same tag, alerts are triggered with suggestion to store the content with same tag in cache.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 9

Predictive Caching System and Method Basing on User Behavior and Chracter Combination Set Trend

Nowadays we are in an era of information explosion. Alongwith the exponential growth of mass users and data in information applications, the importance of Cache is widely known - the data that user requested is stored in Cache, it vastly improves speed to access data for application, and reduces wait time for user, and accordingly it provides better user experiences. But the storage capacity of Cache is limited in real life, it is not able to store all the objects. From this perspective, it should be used to store those objects with higher cache value.

This disclosure is to provide a predictive caching system and method based on user count and behaviour weight in time interval.

Claim point


- Asystem and method to predict and select content into cashingsystem and comprising


- Data defined method for getting weight of a user group with same tag at certain time point;
- Judge system using count and weight curve combination for content selection.

- Adefined tagweight system includinguser profile and behavior.

1


Page 02 of 9

2


Page 03 of 9

3


Page 04 of 9

4


Page 05 of 9


(3) ATagged User Group is comprised of those users who visit the content with same tag. The user count of a Tagged User Group is certainly

5


Page 06 of 9

growing from starting day to this day, but it is changing in time interval(s). For instance, user A visited content with Tag 1 yesterday but not today, if take one day as a time interval, the user group of Tag 1 included user A yesterday but does not includes user A today, accordingly the user count of user group of Tag 1 may decrease today. The user count of individual Tagged User Group in time interval can be collected and counted. Its changing is presented as a curve (please see FIG 4-1). The tag weight of individual Tagged User Group in time interval can be calculated with tag weight of each user (please see formula in FIG 4-2). It is timely updated as the users in a Tagged User Group is changing in time interval and their Tag Weight is changing also. Its changing is presented as a curve (please see FIG 4-1).

6


Page 07 of 9


(4) Amonitoring model is setup to monitor the changing of curve slope. For two curves in (3), the changing of curve slope presents the increasing/decreasing rate of count of user visited a dedicated Content, and presents the increasing/decreasing rate of tagged user group weight (please see FIG 5). If the curve slope reaches a certain threshold and alert is triggered with suggestions to store Content with same tag in cache. These Content are c...