Extended correlation methods in a content transformation engine

摘要

Using a first computer, a content transformation agent is customized for a first source data repository that stores source data. The content transformation agent is provided to the first source data repository. In accordance with the customization, a first stage data feed is constructed in the syndication format by dynamically selecting and transforming within the first source data repository selected data primitives of the source data from the non-syndication format into the first stage data feed in the syndication format. The first stage data feed is sent to an aggregation agent. The aggregation agent receives the first stage data feed from the content transformation agent. Based on specific user criteria, a second stage data feed is generated in a syndication format as a union of selected data primitives from the first stage data feed.

主权项

1. A method for constructing a second stage data feed in a second, neutral format from source data in one or more first non-syndication formats, the method comprising:
customizing, using a computer having a memory and tangible processor, a content transformation agent for a first source data repository that stores the source data; providing the content transformation agent to the first source data repository; in accordance with the customization, constructing a first stage data feed in the second, neutral format by dynamically selecting and transforming within the first source data repository selected data primitives of the source data from the one or more first non-syndication formats into the first stage data feed in the second, neutral format, the first stage data feed comprising a payload of the selected data primitives stored originally in the first source data repository; and sending the first stage data feed to an aggregation agent that generates, based on specific user criteria, a second stage data feed in the second, neutral format as a union of selected data primitives from the first stage data feed, wherein the second stage data feed is generated by correlating data primitives within the first stage data feed using detected relationships among the data primitives and selecting the selected data primitives based on the correlated data primitives.