US-12627747-B2 - Derivation of ontological relevancies among digital content

Abstract

System for ontological evaluation and filtering of digital content evaluates metadata associated with content available from an original content server. The metadata is filtered and evaluated by a processing cluster to develop correlation among content for the formation of content “channels”. In general, the filtering and evaluation criteria use predictive algorithms and seek to identify content that is likely to be desired for download by the consumers located at, for example, a particular multi-dwelling unit. The content, once so correlated, is then grouped or aggregated into “channels”.

Inventors

• Bartow WYATT
• Mark Scifres

Assignees

• Pavlov Media, Inc.

Dates

Publication Date

20260512

Application Date

20200302

Claims (15)

1. 1 . A system for deriving ontological relevancies of digital content to provide the content to a client device connected to a local network without at least some of the delays and latency associated with internet downloads comprising: a metadata processing cluster in a central processing cloud executing on one or more servers for evaluating metadata associated with at least some content available across a network from an original content server, at least in part using predictive algorithms on the result of past usage characteristics of at least some users associated with the central processing cloud, the central processing cloud configured to process, in response to the evaluating step, content that is likely to be downloaded from a remote network location relative to the original content server such that a request to the original content server will result in delivering that content to a user with essentially no latency, and a local content store for storing content that is likely to be downloaded into one or more channels, the local content store located more proximate to a user device than the original content server and connected substantially locally thereto wherein at least a portion of the requested content is delivered to the user device with essentially no latency and essentially at the maximum data rate permitted by the local area network connection.
2. 2 . The system of claim 1 wherein the predictive algorithms develop ontological correlations among the content likely to be downloaded and the past usage characteristics.
3. 3 . The system of claim 1 wherein all personal data associated with the requested content is anonymized.
4. 4 . The system of claim 1 wherein the local content store comprises a network appliance.
5. 5 . The system of claim 1 wherein the local content store comprises a plurality of network appliances.
6. 6 . The system of claim 5 wherein the local content store further comprises a regional appliance.
7. 7 . The system of claim 6 wherein the regional appliance and plurality of network appliances form a tiered network.
8. 8 . The system of claim 1 wherein the local content store comprises a content node.
9. 9 . A method for providing content to a client device connected to a local network at essentially the maximum data rate permitted by the local network connection without at least some of the delays and latency associated with internet downloads from remote content servers comprising the steps of: evaluating in a computer metadata associated with at least some content available across a network from at least one original content server, assessing, in a computer, new requests for content available from the at least one original content server, at a remote server, and in response to the evaluating and assessing steps, applying to the new requests predictive algorithms based at least in part on the result of past usage characteristics of at least some users content that is likely to be downloaded at a network location remote from the original content server such that a request to the original content server will result in delivering that content to a user with essentially no latency, and storing on a local content store, in response to the applying step, selected elements of requested content into one or more channels of aggregated content, where the local content store is located more proximate to a user device than the original content server and connected substantially locally thereto wherein at least a portion of the requested content on the local content store is delivered to the user device with essentially no latency and essentially at the maximum data rate permitted by the local area network connection.
10. 10 . The method of claim 9 further comprising the step of determining whether at least a portion of the requested content already is stored in at least one channel.
11. 11 . The method of claim 9 further comprising the step of maintaining on the local content store some non-channel content.
12. 12 . The method of claim 9 further comprising the step of determining whether the request for content originates from a blacklisted address.
13. 13 . The method of claim 9 further comprising the step of anonymizing the identification of a user making a request for content while at the same time providing information as to the location from which the request originated.
14. 14 . The method of claim 13 wherein the anonymizing occurs at a local appliance rather than at the remote server.
15. 15 . The method of claim 9 wherein destination and host information associated with a request for content is identified from packet and protocol headers.

Description

RELATED APPLICATIONS This application is related to, and claims the benefit of, U.S. Patent Applications Ser. No. 61/770,163, filed Feb. 27, 2013; Ser. No. 61/770,186; Ser. No. 61/770,204; Ser. No. 61/770,211; all filed on Feb. 27, 2013 and incorporated herein by reference as though set forth in full. FIELD OF THE INVENTION The present invention relates generally to the field of internet content delivery systems and methods, and more particularly relates methods, systems and techniques for automatically selecting and caching content at edge servers to provide significantly faster content delivery to, for example, multi-dwelling units. BACKGROUND OF THE INVENTION Fast delivery of internet content has long been desired by users of the internet. However, bandwidth limitations have historically existed which seriously limited the ability of the internet infrastructure to meet the ever-increasing demands of users for more content, delivered more quickly. The limitations of the existing internet infrastructure are nowhere more apparent than the exploding demand for fast downloading of video content from the internet. However, video requires substantially greater bandwidth than most other content and, as a result, such consumer demand has placed substantial strain on the internet infrastructure. While services such as Akamai have attempted to place limited, pre-designated content at the edge of the internet, such efforts are typically limited to icons, images and ads, and do not include the actual content that users desire to see. As a result, there has long been a need for a system which can effectively “speed up” delivery of desired content without substantial delay caused by the inherent bandwidth limitations associated with most internet feeds. SUMMARY OF THE INVENTION The present invention provides an efficient system and method for identifying, caching, and delivering at high speed, the content that users in, for example, a multiple-dwelling unit are likely to want to download based on prior usage characteristics augmented by trend data the system aggregates across a plurality of similar and dissimilar multiple-dwelling units and the users within. In an aspect of the invention, the system comprises local content storage and an associated local network appliance deployed proximate to, or within, a multi-dwelling unit such as an office building, an apartment building, a dormitory, or other business or residential structure. Although, for convenience of illustration, the following description assumes a multi-dwelling unit in many cases, in some embodiments the local appliance can serve either groups of consumers distributed geographically, or only a single consumer. The local network appliance communicates with consumer devices, and also communicates over the internet with original content servers and, importantly in some embodiments, a central processing cloud. The central processing cloud cooperates with the local network appliance to identify content that consumers desire to download. The central processing cloud includes structures and methods for identifying content likely to be downloaded by the consumer devices, and communicates over the internet with the server where the desired content is originally maintained. The metadata associated with the original content is communicated to the central processing cloud, and a copy of the original content along with its associated metadata are downloaded and stored. The metadata is used by a processing cluster, using a variety of filtering and evaluation criteria, to develop ontological correlations, or relevancies, among the content. In general, the filtering and evaluation criteria use predictive algorithms and seek to identify content that is likely to be desired for download by the consumers located at, for example, a particular multi-dwelling unit. The content, once so correlated, is then grouped or aggregated into “channels”. In an embodiment, the content comprising at least one channel is then downloaded to the local network appliance and stored on the local content storage associated with the particular dwelling unit. A high speed local network connects the local network appliance and its associated local content storage, where the channel of content is stored, to the consumer devices. When a consumer at that multi-dwelling unit seeks to download data that is included within the channel, the data is essentially immediately available, without any of the delays and latency associated with conventional internet downloads, and thus the consumer receives the desired content at extremely fast data rates. The central processing cloud includes functionality for identifying the origin of requests for content, using data received from the local network appliance. In at least some embodiments, personal data associated with such requests is “blurred out” or anonymized at the local appliance level, to minimize the risk of personal data being compromised. However, in at