Back to results
US 11,599,544 B2Granted
Primary tagging in a data stream
AssigneeAcxiom Corporation
InventorThomas Lawrence Shrum
Life of this patent
Estimating the term…
Loading patent…
Loading patent…
AssigneeAcxiom Corporation
InventorThomas Lawrence Shrum
Estimating the term…
A method of searching data streams arriving over a network in order to look for a predetermined reserved set of indicia first identifies and then tags the data per the found information in a reserved indicia set. The tags are applied by “StreamTagApps” (STAs), which are small, quickly executing software applications that are each focused on the tagging of data that includes a particular indicium in order to make the identification process rapid. The method thus allows very quick action to be taken in order to maximize the usefulness of the data. Multiple STAs in one or both of horizontal and vertical configurations may be constructed in order to provide any necessary level of speed or take full advantage of an available hardware configuration.
A method for primary tagging of data, comprising the steps of:
A method for primary tagging of data, comprising the steps of:
One or more non-transitory computer readable storage media, having computer-executable instructions embodied thereon, wherein when executed by at least one processor, the computer-executable instructions cause the at least one processor to:
One or more non-transitory computer readable storage media, having computer-executable instructions embodied thereon, wherein when executed by at least one processor, the computer-executable instructions cause the at least one processor to:
An apparatus for primary tagging of data, comprising:
Dependencies are inferred from the claim text.
Drawings are temporarily unavailable. Open the PDF at USPTO
From the US filing history. Foreign family members aren’t listed.
Query execution: data stream processing; Continuous queries
Indexing structures: trees, e.g. B+trees
Query languages: grouping and aggregation
Details of database functions independent of the retrieved data types: clustering; Classification
About 14 years 8 months of term left · ends 15 Jun 2041 (estimated)
Not applicable.
Streams of electronic data are used in many applications today. As the quantity of available data continues to increase in many fields, more advanced methods are required for receiving, storing, and cataloging the data. There is a particularly strong need for new methods of cataloging data in applications where the “freshness” of the data is a significant aspect of the value of the data. For example, data indicating that a particular consumer is “in market” for a new truck must be acted on quickly, or the customer may have already made a purchase from a competitor before the data is fully processed, and thus the opportunity will be missed. Another example is data about a consumer's location, since this may change in an extremely short timeframe. For example, it may be valuable to know that a consumer is standing near a particular retailer or driving toward a particular coffee shop, in which case marketing messages from these retailers may be more effective if they reach the consumer before the opportunity has passed. Thus although these kinds of data may be highly valuable, they become worthless if the data cannot be processed quickly enough to act before the opportunity passes. If one waits to use the data until after all of the processing, writing of data to storage, and formatting and cataloging of data has been complete as in traditional data processing systems, then it may be too late to meaningfully use the data and the value of the data has been lost.
The present invention is directed to a method and system by which an enterprise may scan data streams arriving over a network in order to look for a predetermined reserved set of words, numbers, glyphs, or other indicia, and then tags the data per the found information in the reserved indicia set. The tags are applied by data stream tagging applications or “StreamTagApps” (STAs), which are a set of small, quickly executing software applications that are each focused on the tagging of data that includes a particular word, term, symbol, or other indicia in order to make the identification process as rapid as possible. Because the tags are applied to the incoming data stream, rather than after any subsequent storage and processing steps, information from the tag will be available very quickly, enabling the data processor to act quickly in response to the tags that were applied to the data, rather than waiting for storage, formatting, and full cataloging of the data. In this way, the process and system allows very quick action to be taken in order to maximize the usefulness of the data, in some cases allowing for a real-time response.
FIG. 1 is an exemplary data flow diagram according to an implementation of the invention showing three different data stream tagging applications or “StreamTagApps” (STAs) functioning to tag incoming data.
FIG. 2 is an exemplary data flow diagram according to an implementation of the invention using vertically scaled STAs to tag incoming data.
FIG. 3 is an exemplary data flow diagram according to an implementation of the invention using horizontally scaled STAs to tag incoming data.
Before the present invention is described in further detail, it should be understood that the invention is not limited to the particular embodiments described, and that the terms used in describing the particular embodiments are for the purpose of describing those particular embodiments only, and are not intended to be limiting, since the scope of the present invention will be limited only by the claims.
In the first step of a process according to an implementation of the invention, a variety of data is sourced from many different potential sources. These sources may include the Internet (such as from companies like Amazon, WhatsApp, and Instagram); Internet of Things (IoT) devices such as web cameras, refrigerators, and home automation devices; or from customer data files. As data streams in, data stream tagging applications (“StreamTagApps” or STAs) focused on single data elements (e.g., name, address, account number, or purchased product) peruse the stream for that element alone. The only purpose of each STA is to search for one particular indicium, and thus the coding of the STA can be very short and its operational speed can be very high in order to keep up with the incoming data stream. Upon finding the data element each STA is looking for, it tags the data appropriately and continues to peruse other data in the stream. For example, suppose an item of data in a stream contains both a name and an address. The STA that is looking for names will tag that data item “name” and the STA that is looking for addresses will tag that same data item “address.” These tags can be added as appended text, an appended number or code indicating certain text, or by other means. With enough different STAs sifting through the data stream as it comes in, a great deal of information may be known about the data immediately, without waiting to fully process, store, and catalogue the data. In certain embodiments, this tagging operation can occur while the information is still streaming in from outside of the data processor's firewall, and thus occurs before the data processor executes any other manipulation of the data at all. The tagged data may be sent directly to a marketing process for use of the data in real time, or may be stored in a tagged data database in order to make it acceptable to other processes.
STAs can be scaled in a manner that makes them more efficient. For example, STAs may be scaled vertically by running the STAs on multiple processors or multiprocessor architectures comprising multiple cores. In addition, STAs may be scaled horizontally by having multiple different STAs running simultaneously. Of course even greater efficiency can be achieved by scaling the STAs both vertically and horizontally, thereby handling data volumes of any size efficiently by scaling up the STAs in a manner corresponding with the data volume.