System and method for parallel urgency-based data analysis

Abstract

A global architecture (GLP), as disclosed herein, is based on the thin server architectural pattern; it delivers all its services in the form of web services and there are no user interface components executed on the GLP. Each web service exposed by the GLP is stateless, which allows the GLP to be highly scalable. The GLP is further decomposed into components. Each component is a microservice, making the overall architecture fully decoupled. Each microservice has fail-over nodes and can scale up on demand. This means the GLP has no single point of failure, making the platform both highly scalable and available. The GLP architecture provides the capability to build and deploy a microservice instance for each course-recipient-user combination. Because each student interacts with their own microservice, this makes the GLP scale up to the limit of cloud resources available—i.e. near infinity.

What is claimed is: 
     
         1 . A system for redundant content communications, the system comprising:
 a memory comprising:
 an asset database comprising a plurality of package-data assets; and 
 a model database comprising a plurality of statistical models; 
   at least one server comprising: a batch layer; and a speed layer, wherein the at least one server is configured to:
 receive an event stream comprising event data indicative of user actions; 
 classify event data within the event stream as either for immediate processing or for batch processing; 
 process event data; 
 receive a content request; 
 identify data relevant to the content request, wherein the data relevant to the content request comprises outputs of the speed layer and outputs of the batch layer, wherein the output of the batch layer comprises a user model; 
 select a package-data asset based on the data received from both the speed layer and the batch layer; and 
 provide the selected package-data asset. 
   
     
     
         2 . The system of  claim 1 , wherein event data is classified by a data ingestion interface. 
     
     
         3 . The system of  claim 2 , wherein the data ingestion interface formats the event data for immediate processing and the event data for batch processing. 
     
     
         4 . The system of  claim 1 , wherein classifying event data comprises:
 identifying, from within the event stream, event data for immediate processing; and   generating, from the event stream data a first set of data for batch processing.   
     
     
         5 . The system of  claim 4 , wherein classifying event data further comprises
 identifying from within the event stream, linked events; and   grouping linked events.   
     
     
         6 . The system of  claim 4 , wherein processing event data comprises:
 processing of the events for immediate processing with the speed layer; and   storing the first data set for batch processing.   
     
     
         7 . The system of  claim 1 , wherein the data indicative of user actions comprises data indicative of all user interactions with a user device. 
     
     
         8 . The system of  claim 1 , wherein the event stream is received from an activity stream API. 
     
     
         9 . The system of  claim 1 , wherein the user model identifies a mastery level of a user associated with the content request. 
     
     
         10 . The system of  claim 1 , wherein the output of the batch layer is provided before performing of batch processing of the received event data classified for batch processing. 
     
     
         11 . A method of hybrid event processing, the method comprising:
 receiving an event stream comprising event data indicative of user actions;   classifying event data within the event stream as either for immediate processing or for batch processing;   processing event data;   receiving a content request;   identifying data relevant to the content request, wherein the data relevant to the content request comprises outputs of a speed layer and outputs of a batch layer, wherein the output of the batch layer comprises a user model;   selecting a package-data asset based on the data received from both the speed layer and the batch layer; and   providing the selected package-data asset.   
     
     
         12 . The method of  claim 11 , wherein event data is classified by a data ingestion interface. 
     
     
         13 . The method of  claim 12 , wherein the data ingestion interface formats the event data for immediate processing and the event data for batch processing. 
     
     
         14 . The method of  claim 11 , wherein classifying event data comprises:
 identifying, from within the event stream, event data for immediate processing; and   generating, from the event stream a first set of data for batch processing.   
     
     
         15 . The method of  claim 14 , wherein classifying event data further comprises
 identifying from within the event stream, linked events; and   grouping linked events.   
     
     
         16 . The method of  claim 14 , wherein processing event data comprises:
 processing of the events for immediate processing with the speed layer; and   storing the first data set for batch processing.   
     
     
         17 . The method of  claim 11 , wherein the event data indicative of user actions comprises data indicative of all user interactions with a user device. 
     
     
         18 . The method of  claim 11 , wherein the event stream is received from an activity stream API. 
     
     
         19 . The method of  claim 11 , wherein the user model identifies a mastery level of a user associated with the content request. 
     
     
         20 . The method of  claim 11 , wherein the output of the batch layer is provided before performing of batch processing of the received event data classified for batch processing.