DATA REFINING ENGINE FOR HIGH PERFORMANCE ANALYSIS SYSTEM AND METHOD
Price and product attributes from webpages are analyzed over time to identify price changes specific to products on individual webpages and for products across all webpages as well as to identify longitudinal correlations between price changes and product attributes. Users may search the data and set alerts.
This application claims the benefit of and incorporates by reference U.S. Provisional Patent Application No. 61/675,492, filed on Jul. 25, 2012. This application also incorporates by reference co-pending U.S. patent application Ser. No. ______, filed on Jul. 25, 2013, titled, “Adaptive Gathering of Structured and Unstructured Data System and Method,” which application also claims the benefit of U.S. Provisional Patent Application No. 61/675,492.
FIELDThis disclosure relates to a method and system to analyze price and product information.
BACKGROUNDThe following description includes information that may be useful in understanding the present invention. It is not an admission that any of the information provided herein is prior art or relevant to the presently claimed invention, or that any publication specifically or implicitly referenced is prior art.
Search engines, such as Google, Bing, and others search and index vast quantities of information on the Internet. “Crawlers” (a.k.a. “spiders”) utilize URLs obtained from a “queue” to obtain content, usually from web pages. The crawlers or other software store and index some of the content. Users can then search the indexed content, view results, and follow hyperlinks back to the original source or to the stored content (the stored content often being referred to as a “cache”). Computing resources to crawl and index, however, are not limitless. The URL queues are commonly prioritized to direct crawler resources to web page servers which can accommodate the traffic, which do not block crawlers (such as according to “robots.txt” files commonly available from webpage servers), which experience greater traffic from users, and which experience more change in content.
Conventional search engines, however, are not focused on price and product information. If a price changes on a webpage, but the rest of the webpage remains the same, traditional crawlers (or the queue manager) will not prioritize the webpage position in the queue, generally because the price is a tiny fraction of the overall content and the change is not labeled as being significant; conversely, if the webpage changes, but the price and/or product information remains the same, the change in webpage content may cause a traditional crawler to prioritize the webpage position in the queue due to the overall change in content, notwithstanding that that price and product information remained the same.
Conventional search engines, if presented with a query, will find corresponding products. For example, it is possible to search for “men's shoes” and to then be presented with a webpage comprising search results for hundreds of thousands of webpages for men's shoes. The search result may further be narrowed by category of men's shoes, brand, and store. Search engines have been incorporated into online stores, wherein a user may search for products, by keyword and/or by category and results can be ordered by price.
Price history, however, is only narrowly viewed and, when it is, never in the context of a rich attribute set which explores, in detail, which attributes are associated with changes in price. Price histories are not made available in real time, and do not allow intricate comparisons based on stores, merchants, brands, regions, time/date, and other dimensions.
The following Detailed Description provides specific details for an understanding of various examples of the technology. One skilled in the art will understand that the technology may be practiced without many of these details. In some instances, structures and functions have not been shown or described in detail or at all to avoid unnecessarily obscuring the description of the examples of the technology. It is intended that the terminology used in the description presented below be interpreted in its broadest reasonable manner, even though it is being used in conjunction with a detailed description of certain examples of the technology. Although certain terms may be emphasized below, any terminology intended to be interpreted in any restricted manner will be overtly and specifically defined as such in this Detailed Description section.
Unless the context clearly requires otherwise, throughout the description and the claims, the words “comprise,” “comprising,” and the like are to be construed in an inclusive sense, as opposed to an exclusive or exhaustive sense; that is to say, in the sense of “including, but not limited to.” As used herein, the term “connected,” “coupled,” or any variant thereof means any connection or coupling, either direct or indirect between two or more elements; the coupling of connection between the elements can be physical, logical, or a combination thereof. Additionally, the words, “herein,” “above,” “below,” and words of similar import, when used in this application, shall refer to this application as a whole and not to particular portions of this application. When the context permits, words using the singular may also include the plural while words using the plural may also include the singular. The word “or,” in reference to a list of two or more items, covers all of the following interpretations of the word: any of the items in the list, all of the items in the list, and any combination of one or more of the items in the list.
Certain elements appear in various of the Figures with the same capitalized element text, but a different element number. When referred to herein with the capitalized element text but with no element number, these references should be understood to be largely equivalent and to refer to any of the elements with the same capitalized element text, though potentially with differences based on the computing device within which the various embodiments of the element appears.
As used herein, a Uniform Resource Identifier (“URI”) is a string of characters used to identify a resource on a computing device and/or a network, such as the Internet. Such identification enables interaction with representations of the resource using specific protocols. “Schemes” specifying a syntax and associated protocols define each URI.
The generic syntax for URI schemes is defined in Request for Comments (“RFC”) memorandum 3986 published by the Internet Engineering Task Force (“IETF”). According to RFC 3986, a URI (including a URL) consists of four parts:
-
- <scheme name>: <hierarchical part> [?<query>] [#<fragment>]
A URI begins with a scheme name that refers to a specification for assigning identifiers within that scheme. The scheme name consists of a letter followed by any combination of letters, digits, and the plus (“+”), period (“.”), or hyphen (“-”) characters; and is terminated by a colon (“:”).
The hierarchical portion of the URI is intended to hold identification information that is hierarchical in nature. Often this part is delineated with a double forward slash (“//”), followed by an optional authority part and an optional path.
The optional authority part holds an optional user information part (not shown) terminated with “@” (e.g. username:password@), a hostname (i.e., domain name or IP address, here “example.com”), and an optional port number preceded by a colon “:”.
The path part is a sequence of one or more segments (conceptually similar to directories, though not necessarily representing them) separated by a forward slash (“/”). If a URI includes an authority part, then the path part may be empty.
The optional query portion is delineated with a question mark and contains additional identification information that is not necessarily hierarchical in nature. Together, the path part and the query portion identify a resource within the scope of the URI's scheme and authority.
The query string syntax is not generically defined, but is commonly organized as a sequence of zero or more <key>=<value> pairs separated by a semicolon or ampersand, for example:
-
- key1=value1;key2=value2;key3=value3 (Semicolon), or
- key1=value1&key2=value2&key3=value3 (Ampersand)
Much of the above information is taken from RFC 3986, which provides additional information related to the syntax and structure of URIs. RFC 3986 is hereby incorporated by reference, for all purposes.
As used herein, “Product” shall be understood to mean “products or services.” References to “Product Attribute” herein shall be understood to mean “product or service attribute.” As used herein, “Products” are associated with iPIDs.
As used herein, an “iPID” or iPID 330 is a unique identifier assigned within the Indix System to a URI for a product, such as URI 305. The iPID 330 may be, for example, a hash of URI 305. When multiple URIs 305 from a common base domain name lead to webpages which, when parsed for Price and Product Attributes and product the same Parse Result 325 (notwithstanding that the webpages may contain other Content which does not contribute to the Parse Result 325) may be labeled as equivalent in, for example, the Equivalent iPID 334 record and may be treated as the same iPID 330.
As used herein, a “Master iPID” or “MPID” or MPID 332 is an iPID 330 assigned to a group of iPIDs 330 derived from URIs 305 which lead to webpages offering the same Product for sale. An MPID is generally meant to identify a single Product, generally produced by a common manufacturer, though the Product may be distributed and sold by multiple parties.
iPIDs and MPIDSs are associated with Price Attribute 340 records and Product Attribute 345 records.
A Price Attribute 340 record may comprise one or more records comprising, for example, values which encode an iPRID which may be an identifier for a price observed at a particular time, an iPID (discussed above), a Product Name (a “Product Name” value in this record may also be referred to herein as a “Product”), a Standard Price, a Sale, a Price, a Rebate amount, a Price Instructions record (containing special instructions relating to a price, such as that the price only applies to students), a Currency Type, a Date and Time Stamp, a Tax record, a Shipping record (indicating costs relating to shipping to different locations, whether tax is calculated on shipping costs, etc.), a Price Validity Start Date, a Price Validity End Date, a Quantity, a Unit of Measure Type, a Unit of Measure Value, a Merchant Name (with the name of a merchant from whom the Product is available; a “Merchant Name” value in this record may also be referred to herein as a “Merchant”), a Store Name (a Merchant may have multiple stores; a “Store Name” value in this record may also be referred to herein as a “Store”), a User ID, a Data Channel (indicating the source of the Price Attribute 340 record, such as an online crawl, a crowdsource, a licensed supplier of price information, or from a merchant), a Source Details record (for example, indicating a URI, a newspaper advertisement), an Availability Flag, a Promotion Code, a Bundle Details record (indicating products which are part of a bundle), a Condition Type record (indicating new, used, poor, good, and similar), a Social Rank record (indicating a rank of “likes” and similar of the price), a Votes/Likes record (indicating a number of “likes” and similar which a Price or Product has received), a Price Rank record, a Visibility Indicator record (indicating whether the price is visible to the public, whether it is only visible to a Merchant, or the like), a Supply Chain Reference record (indicating whether the price was obtained from a retailer, a wholesaler, or another party in a supply chain), a Sale Location (indicating a geographic location where the product is available at the price), a Manufactured Location record (indicating where the product was produced or manufactured), a Launch Date record (indicating how long the product has been on the market), and an Age of Product record (indicating how long the product was used by the user). When capitalized herein, the foregoing terms (such as Product, Price, Merchant, Store, Source Details, etc.) are meant to refer to values in a Price Attribute 340 record.
A Product Attribute 345 record may comprise, for example, values encoding features of or describing a Product. The entire Product Attribute 345 schema may comprise thousands of columns, though only tens or hundreds of the columns may be applicable to any given Product. An example set of values in a Product Attribute 345 record for a ring is as follows: Title, “Sterling Silver Diamond & Blue Topaz Ring;” Brand, “Blue Nile;” Category (such as, for example, a Category 335 in a category schema), “rings;” Metal Name, “silver;” Stone Shape, “cushion;” Stone Name, “topaz;” Width, “3 mm;” Stone Color, “blue;” Product Type, “rings,” Birthstone, “September;” and Setting Type, “prong.” An example set of Product Attributes 345 for a shoe is as follows: Brand, “Asics;” Category (such as, for example, a Category 335 in a category schema or taxonomy), “Men's Sneakers & Athletic;” Shoe Size, “8;” Product Type, “wrestling shoes,” Color, “black;” Shoe Style, “sneakers;” Sports, “athletic;” Upper Material, “mesh.” When capitalized herein, the foregoing terms (such as Brand, Category, Metal Name, Product Type, etc.) are meant to refer to values in a Product Attribute 345 record.
As used herein, “Content” comprises text, graphics, images (including still and video images), audio, graphical arrangement, and instructions for graphical arrangement, including HTML and CSS instructions which may, for example, be interpreted by browser applications.
As used herein, “Event” is information generally in news or current events. Events may be found in Content. Listing Pages, Product Pages, and Event Pages are all examples of Webpage Types 350.
As used herein, “PriceDNA” comprises a Product Attribute 345 record, one or more Price Attribute 340 records, the output of the Core Price Routine 500 (generally found in the Core Price 380 records), and the output of the Insights 600 routine (generally found in the Insights 375 records).
As used herein, a “Brand” is a family or group of Products sold by or under a common trademark, such as the “Nike®” Brand, which sells under this trademark a family of shoes, exercise equipment, and other apparel. Brand is a value within a Product Attribute 345 record.
As used herein, a “Store” is an online or physical sales venue. A Store is a value within a Price Attribute 340 record.
As used herein, a “Merchant” is an operator of one or more Stores. A Merchant is a value in a Price Attribute 340 record.
Generally, an Analysis Routine 400 obtains Price Attribute 340 and Product Attribute 345 records from the Indix Database 300 shortly after the records are produced following a crawl of webpages accessed via the URIs 305. The Analysis Routine 400 merges the records, performs a Core Price Routine 500 to develop core price information, such as changes in price, and exports the records and the result to the Core Price Routine 500 to a sequential file which is indexed. The result of the Core Price Routine 500 may be searched and accessed by users in close to real-time. The Analysis Routine 400 also performs an Insight Routine 600. The Insight Routine 600 comprises a set of sub-routines for deriving additional information from the Price Attribute 340 and Product Attribute 345 records and from the output of the Core Price Routine 500. Generally, the Insight Routine 600 identifies what Product Attributes 345 and Price Attributes 340 across the datasets are associated with the changes in price. The output of the Insight Routine 600 is also stored in the Indix Database 300 and may be searched and accessed by users, though the accessible values may be refreshed more slowly than the data from the Core Price Routine 500. A User Contact Routine 1700 allows users to search and obtain information and to set alerts relative to the information in the Indix Database 300.
Also illustrated in
Also illustrated in
Also illustrated in
Also illustrated in
Interaction among devices illustrated in
In
The Network 150 comprises computers, network connections among the computers, and software routines to enable communication between the computers over the network connections. Examples of the Network 150 comprise an Ethernet network, the Internet, and/or a wireless network, such as a GSM, TDMA, CDMA, EDGE, HSPA, LTE or other network provided by a wireless service provider, or a television broadcast facility. Connection to the Network 150 may be via a Wi-Fi connection. More than one network may be involved in a communication session between the illustrated devices. Connection to the Network 150 may require that the computers execute software routines which enable, for example, the seven layers of the OSI model of computer networking or equivalent in a wireless phone network.
This paper may discuss a first computer as connecting to a second computer (such as a Crawl Agent 400 connecting to the Indix Server 200) or to a corresponding datastore (such as to Indix Database 300); it should be understood that such connections may be to, through, or via the other of the two components (for example, a statement that a computing device connects with or sends data to the Indix Server 200 should be understood as saying that the computing device may connect with or send data to the Indix Database 300). References herein to “database” should be understood as equivalent to “datastore.” Although illustrated as components integrated in one physical unit, the computers and databases may be provided by common (or separate) physical hardware and common (or separate) logic processors and memory components. Though discussed as occurring within one computing device, the software routines and data groups used by the software routines may be stored and/or executed remotely relative to any of the computers through, for example, application virtualization.
The Indix Server Memory 250 stores program code for software routines, such as, for example, Analysis Routine 400, Core Price Routine 500, Insight Routine 600, Volatility Routine 700, Substitution Routine 800, Mix Routine 900, Prediction Routine 1000, Competition Routine 1100, Promotion Routine 1200, Leadership Routine 1300, Premium Routine 1400, Price Range Routine 1500, Reach Routine 1600, and User Contact Routine 1700 as well as, for example, browser, email client and server routines, client applications, and database applications (discussed further below). Additional data groups for routines, such as for a webserver and web browser, may also be present on and executed by the Indix Server 200 and the other computers illustrated in
In addition, the Indix Server Memory 250 also stores an Operating System 255. These software components may be loaded from a non-transient Computer Readable Storage Medium 295 into Indix Server Memory 250 of the computing device using a drive mechanism (not shown) associated with a non-transient Computer Readable Storage Medium 295, such as a floppy disc, tape, DVD/CD-ROM drive, memory card, or other like storage medium. In some embodiments, software components may also or instead be loaded via a mechanism other than a drive mechanism and Computer Readable Storage Medium 295 (e.g., via Network Interface 230).
The computing device 200 may also comprise hardware supporting input modalities, Input 245, such as, for example, a touchscreen, a camera, a keyboard, a mouse, a trackball, a stylus, motion detectors, and a microphone. The Input 245 may also serve as a Display 240, as in the case of a touchscreen display which also serves as Input 245, and which may respond to input in the form of contact by a finger or stylus with the surface of the Input 245.
The computing device 200 may also comprise or communicate via Bus 220 with Indix Datastore 300, illustrated further in
At box 500, the Analytic Routine 400 performs the Core Price Routine 500 (discussed further below). At box 410, for each iPID 330 associated with a Price Attribute 340 or Product Attribute 345 record in box 405, the Analytic Routine 400 appends the then-current Price Attribute 340 record (of box 405) to a set of Price Attribute 340 records associated with each iPID 330 (each iPID 330 may be associated with a set of Price Attribute 340 records). At box 415, for each iPID 330 associated with a Price Attribute 340 or Product Attribute 345 record in box 405, the Analytic Routine 400 merges the then-current Product Attribute 345 record (of box 405) into a Product Attribute 345 record associated with each iPID 330 (each iPID 330 may be associated with one Product Attribute 340 record). In this merger, new values overwrite old values unless the old record is longer or unless the old record otherwise is judged to be of higher quality (such as if the old record uses fewer words, but the words are less common than the words in the new record); if a new record does not have a value where an old value exists, the old value may be left.
At box 420, the output of the Core Price Routine 500 and of boxes 410 and 415 are output to a Sequential File 365 record, which Sequential File 365 record is stored, for example, in the Indix Database 300 and which Sequential File 365 is indexed, for example, to allow the contents of the Sequential File 365 to be searched and values in it accessed as it and the index are updated. Updates may occur, for example, in close-to real-time, following crawl of a webpage and output of new Price Attribute 340 and Product Attribute 345 records.
At box 600, the Analytic Routine 400 performs the Insight Routine 600 utilizing and expanding upon the output of the Core Price Routine 500 and the boxes, above. Generally, the Insight Routine 600 identifies what Product Attributes 345and Price Attributes 340 across the datasets are associated with the changes in price. At box 435 the Analytic Routine 400 stores the output of the Insight Routine 600 in the Indix Database 300 as Insights 375. At box 1700, the Analytic Routine 400 performs the User Contact Routine 1700. Utilizing the User Contact Routine 1700, users may query the Indix Database 300 and set alerts.
At box 520, an MPID 332 associated with the iPID 330 may be obtained. At box 525, the high, low, average, mean, magnitude, and number of price values over several time periods may be calculated for the MPID 332 utilizing the new value associated with the iPID 330 from box 515. The iPID 330 may be a hash of a URI 305 and the result of box 515 is thus limited to a particular sales channel (typically a Store or Merchant) for a particular Product (taking into account that duplicate iPIDs 330 from a base domain name may be treated as equivalent); the MPID 332 is assigned to all iPIDS 330 which represent the same Product, so the MPID version of this calculation in box 525 returns values relating to the Product across Stores, Merchants, Locations, etc. The calculation of box 525 may return values which are or may be sorted by, for example, Store, Merchant, Location (such as Region), and by time periods such as a Season. The output may be saved, for example, to the Core Price 380 records, and indexed.
At box 535, all calculations and other routines which utilize the values for the iPID 330 and the associated MPID 332 may insert the new values calculated for the iPID 330 and the MPID 332 and may recalculate the values. For example, the high, low, average, mean, magnitude, and number of price changes over time periods by Category 335, such as a Category 335 associated with the iPID 330, may be calculated. The output may be saved, for example, to the Core Price 380 records, and indexed.
Calculations or other routines which utilize the values calculated in
At box 599, the Core Price Routine 500 may return, for example, to the Analysis Routine 400.
The Insights Routine 700 may perform one or more of a set of sub-routines. At box 700, a Volatility Routine 700 may be performed to determine the volatility of prices relative to the many dimensions available in the PriceDNA. At box 800, a Substitution Routine 800 determines substitutes for an iPID 330, MPID 332, or Category 335. At box 900, a Mix Routine 900 determines “how many” relative to the many dimensions available in the PriceDNA. At box 1000, a Prediction Routine 1000 makes price predictions relative to the many dimensions available in the PriceDNA. At box 1100, a Competition Routine 1100 determines competitors relative to a Product, Store, or Brand. At box 1200, a Promotion Routine 1200 determines promotions relative to Products, Stores, Brand, Seasons, and other dimensions available in the PriceDNA. At box 1300, a Leadership Routine 1300 determines which Products lead or follow others in terms of price changes. At box 1400, a Premium Routine 1400 determines which Products in a Category 335 charge higher (premium) prices. At box 1500, a Price Range Routine 1500 determines the number of price ranges and maximum and minimum for iPIDs, MPIDs, and categories. At box 1600, a Reach Routine 1600 determines the reach of an iPID or MPID in terms of the number of people who visit a sales venue.
At box 720, the benchmark number of Price changes in the period of time may be determined. The benchmark may be, for example, the VBF relative to additional criteria, such as, for example, the VBF for a Product (or MPID), plus 1, divided by the maximum VBF of other Products in the same Category as the Product, multiplied by 100 over 101. The benchmark VBF for a Category may be determined by the VBF for the Category, plus 1, divided by the maximum VBF of the Category, multiplied by 100 over 101. The benchmark VBF for a Merchant may be the VBF of the Merchant, plus 1, divided by the maximum VBF of the Merchant, multiplied by 100 over 101. The benchmark VBF for a Brand may be the VBF of the Brand, plus 1, divided by the maximum VBF of the Brand, multiplied by 100 over 101. The values may be saved to the Insights 375 records.
At box 820, the iPIDs 330 in the Category of box 805 and with a Price value within the Price Band of box 810 are identified, such as from the Core Price 380 records. At box 825, the result of box 820 may be subdivided, grouped, or filtered by Region, Time, Used/New, and according to other dimensions available in the PriceDNA. At box 830 the Substitution Routine 800 may iterate over the remaining iPIDs 330 in the Category 335. At box 835, the results may be saved as Substitutes, such as to the Insights 375 records. At box 839, the process may return.
In a second example of an embodiment of a Substitution Routine 800 illustrated in
Boxes 860 through 870 may iterate for each iPID 330 within the MPIDs 332 and Attribute 345 match of box 850. At box 865, the iPIDs 330 with a Price value within the Price Band of box 855 and with the Product Attribute 345 match or overlap of box 850 are identified. The result of box 865 may be subdivided or grouped further by sub-Price Ranges. At box 870 the Substitution Routine 800 may iterate over the remaining iPIDs 330 in the MPIDs 332 within the Category 335. At box 871, the results may be saved as Substitutes in the Insights 375 records. At box 874, the process may return.
In a third example of an embodiment of a Substitution Routine 800 illustrated in
Boxes 890 through 897 may iterate for each iPID 330 within the MPIDs 332 and Attribute 345 match of box 885. At box 895, the iPIDs 330 with the Product Attribute 345 match or overlap of box 885 and in the bottom of a Price Range or Price Band relative to the starting iPID 330 are identified. At box 896 the top or bottom five (or another subset) of box 895 may be selected. At box 897 this embodiment of the Substitution Routine 800 may iterate over the remaining iPIDs 330 in the MPIDs 332 within the Category 335. At box 898, the results may be saved as Substitutes in the Insights 375 records. At box 899, the process may return.
At block 1020, to the first parameter may be added the second parameter multiplied by the last price of the Product from block 1010. At block 1025 an error term may be added to the result of block 1020. At block 1030 a confidence interval may be calculated. At block 1035 the result may be saved as Predictions in the Insights 375 records. At block 1035 the Prediction Routine 1000 may then return.
In
In this formula, the parameters of the model are estimated using the original least squares method as follows:
At box 1115, a determination may be made regarding whether or not the entities of box 1105 have 70% or more overlapping Products, per the Products of box 1110. The affirmative output of this box may be saved as Competitors in the Insights 375 records.
At box 1120, the Competitors may be filtered by, for example, on or more of Store, Substitute, Substitute by Price Band, Brand, Location (including Region), Time (including Season), and whether the Products are sold as used or new. Which criteria are used in the filter may be determined by input from a user. The output of box 1120 may be saved in the Insights 375 records.
At box 1125, the average price of Products in the Category 335 of box 1105 may be obtained relative to, for example, the Category 335, Substitute, Substitute by Price Band, Brand, Location, Time, used/new status, and other criteria. At box 1130, the output of box 1125 may be ranked and saved as Price Competitiveness in the Insights 375 records.
At box 1135, a Store and Location for a target Product may be obtained, such as from a user. At box 1145, the Competitors from box 1115 may be obtained or determined and the Competitors filtered to select only Competitors with sales in the Location of box 1135. At box 1145, Stores in the Location which are the same as the Store of box 1135 may be removed from the set of Competitors, leaving the remainder (those not removed).
At box 1150, the output of box 1150 may be placed in a Voroni Diagram or similar data structure, with the location in the Vononi Diagram being based on physical location of the Stores of the Competitors. Generally, a Voroni Diagram determines the distance between objects in a geometric manner, rather than a power-law manner. At box 1155, the distance between the target Store and each Competitor may be ranked. At box 1160, the output of box 1160 may be saved as Reach Competitiveness in the Insights 375 records.
At box 1215, the number, length, date/time, and magnitude of the Promotions may be determined and saved as Promotions in the Insights 375 records. Alternatively, the number, length, date/time, and magnitude of the low-points in the graph of box 1210 may be determined and saved as Promotions in the Insights 375 records. At box 1220, the output of box 1215 may be filtered by criteria such as, for example, date/time, Price Band, Location (including Region), Season, and Holidays. The criteria may be received from, for example, a user and/or a default set of criteria may be applied, with the result of each being saved in the Insights 375 records.
At box 1225 a time period and a Merchant may be obtained, such as from a user; the Merchant may be associated with the Product of box 1205. At box 1225, the number of Products sold by the Merchant in Promotion during the time period may be determined.
At box 1230, the result of box 1215 may be benchmarked relative to average Promotion times, durations, and magnitude for other Products (including other Bundles of the Product), the timing of Promotions for other Products, relative to the magnitude of Promotions for other Products, relative to the Products associated with a Brand, relative to all Products sold at a Store, relative to Products in a Price Band, and relative to Competitors and Substitutes. The result may be saved in the Insights 375 records.
At box 1320, the Price deltas determined at box 1315 may be matched, to determine if any of the Price deltas with the same absolute value (positive or negative) occurred within a time window of one another (deltas beyond the time window may not be considered to be correlated), with the result being saved as a Leader/Follower indication in the Insights 375 records.
At box 1325, the matching Price deltas of box 1320 may be graphed according to time. At box 1330, the result of box 1325 may be filtered by criteria such as Region, Rime, Date/Time, Season, Price Band, and Store.
At box 1335, the number of Leaders and Followers may be determined relative to a time period. At box 1340, the average lead/follow time may be determined. At box 1345, leaders/followers with respect to exact Product matches (for different Stores selling the same Product, determined at box 1330) may be identified. At box 1350, the results may be benchmarked relative to the number of leaders/followers and other criteria. The result of various of the boxes in
At box 1425, the Product Attributes 345 of the Products and Substitutes of box 1410 may be obtained and clustered by similarity. At box 1430, the Product Attributes 345 unique to or dominant in the Premium Products, determined by the clusters of box 1425, may be identified and saved in the Insights 375 records.
At box 1435, user votes regarding Product Attributes 345 of Premium Products may be received. At box 1440, the user votes may be tallied and, at box 1445, the “winning” Product Attributes 345 (with the most votes) may be set as the Product Attributes 345 associated with the Premium Products in the Insights 375 records.
At box 1520, the Channel Range for the Product may be set as the minimum and maximum of the Prices of box 1510 and saved in the Insights 375 records. At box 1525, the results of boxes 1510, 1515, and 1520 may be filtered by, for example, Region, Date/Time, and according to other criteria and saved in the Insights 375 records. At box 1530, the Price Ranges may be determined relative to all Products in a Category 335, all Products by a Brand, and relative to a benchmark which may be, for example, the maximum number of Price Ranges within a Category 335. The result thereof may be saved as Price Ranges in the Insights 375 records.
The above Detailed Description of embodiments is not intended to be exhaustive or to limit the disclosure to the precise form disclosed above. While specific embodiments of, and examples are described above for illustrative purposes, various equivalent modifications are possible within the scope of the system, as those skilled in the art will recognize. For example, while processes or blocks are presented in a given order, alternative embodiments may perform routines having operations, or employ systems having blocks, in a different order, and some processes or blocks may be deleted, moved, added, subdivided, combined, and/or modified. While processes or blocks are at times shown as being performed in series, these processes or blocks may instead be performed in parallel, or may be performed at different times. Further, any specific numbers noted herein are only examples; alternative implementations may employ differing values or ranges.
Claims
1. A computer implemented method of processing information from webpages, the method comprising:
- receiving a first and a second set of price and product attributes for a first product, which attributes comprise: a first identifier of a first identifier-type derived from a URI which links to a webpage offering the product for sale, a second identifier of a second identifier-type assigned to all instances of the product as offered for sale at any URI, and a first category in a category taxonomy;
- performing a first URI-specific price analysis of price values in the first and second sets of price and product attributes to identify changes in price for the first product and associating the result with the first identifier of the first identifier-type and saving the result as a first URI-specific core price result;
- receiving a third and fourth set of price and product attributes for a second product, which attributes comprise: a third identifier of the first identifier-type, a fourth identifier of the second identifier-type, and a second category in the category taxonomy;
- performing a second URI-specific price analysis of price values in the third and fourth sets of price and product attributes to identify changes in price for the second product and associating the result with the third identifier of the first identifier-type and saving the result as a second URI-specific core price result;
- when the second identifier and the fourth identifiers are the same, performing a first non-URI-specific price analysis utilizing the first and second URI-specific core price results to identify changes in price according to the second identifier-type and saving the result as a first non-URI-specific core price result;
- saving and indexing the output of the URI-specific and non-URI-specific price analyses in a first file structure and making the first file structure available to be searched substantially as the sets of product and price attributes are received;
- performing a meta-analysis utilizing the URI-specific and non-URI-specific core price results to identify what product and price attributes across the datasets are associated with the changes in price; and
- saving and indexing the output of the meta-analysis as a second file structure and making the second file structure available to be searched.
2. The method of claim 1, further comprising merging new product attribute records into prior product attribute records and saving new price attribute records along with prior price attribute records.
3. The method of claim 1, wherein the URI-specific price analysis comprises determining the high, low, average, mean, magnitude and number of price changes over at least one time period for the price and product attributes associated with the same identifier of the first identifier-type.
4. The method of claim 1, wherein the non-URI-specific price analysis comprises determining the high, low, average, mean, magnitude and number of price changes over at least one time period for the price and product attributes associated the same identifier of the second identifier-type.
5. The method of claim 4, wherein at least one of the first and second identifier-types are further associated with at least one of a store, a merchant, and a location and wherein the non-URI-specific price analysis produces results associated therewith.
6. The method of claim 1, wherein the price attributes comprise at least one of a time, a product name, a price, a quantity, a unit of measurement, a merchant name, a store name, a bundle detail, and a location.
7. The method of claim 1, wherein the product attributes comprise at least one of a title, a brand, a category in the category taxonomy, a color, a product type, and a size.
8. The method of claim 1, further comprising receiving a user query and executing the query relative to the first and/or second file structures.
9. The method of claim 1, further comprising receiving a user query, a schedule for executing the query, executing the query at the scheduled time on the first and/or second file structures, and alerting the user regarding the result of the query.
10. The method of claim 1, wherein the first and second file structures may be searched by at least one of the first identifier-type, the second identifier-types, or a category in the category taxonomy.
11. The method of claim 1, wherein the first and second categories are the same.
12. The method of claim 1, wherein the meta-analysis determines the volatility of price changes over time for each of the first and second products.
13. The method of claim 12, wherein the volatility is determined by counting the number of price changes in a time period according to at least one of the first identifier-type, the second identifier-type, a brand, a region, a price band, and a category in the category taxonomy.
14. The method of claim 1, wherein the meta-analysis determines whether one of the products is a substitute for the other.
15. The method of claim 14, wherein whether one of the products is a substitute for the other is determined by determining if the first and second products are in the same category in the category taxonomy and by determining whether the first and second products are within a price band within the category.
16. The method of claim 15, further comprising determining if the first and second products share at least fifty-percent of the same product attributes.
17. The method of claim 1, wherein the meta-analysis determines predictions regarding the future prices for the products.
18. The method of claim 17, wherein the predictions are determined by obtaining the last price of at least one of the products from the URI-specific core price associated therewth, calculating or obtaining first and second linear regression parameters, multiplying the second linear regression parameter by the last price and adding this to the first linear regression parameter.
19. The method of claim 1, wherein the price and product attributes comprise at least one of a store, merchant, or brand and the meta-analysis determines products associated therewith and competitors thereof.
20. The method of claim 1, wherein the meta-analysis determines whether a price change for the first product leads or follows a price change for the second product.
21. The method of claim 1, wherein the meta-analysis determines whether the first or second product is a premium product relative to the other.
22. The method of claim 1, wherein the meta-analysis determines the price ranges in which the products are offered for sale.
23. A webpage information processing computing apparatus, the apparatus comprising a processor and a memory storing instructions that, when executed by the processor, configure the apparatus to: save and index the output of the meta-analysis as a second file structure and make the second file structure available to be searched.
- receive a first and a second set of price and product attributes for a first product, which attributes comprise: a first identifier of a first identifier-type derived from a URI which links to a webpage offering the product for sale, a second identifier of a second identifier-type assigned to all instances of the product as offered for sale at any URI, and a first category in a category taxonomy;
- perform a first URI-specific price analysis of price values in the first and second sets of price and product attributes to identify changes in price for the first product and associating the result with the first identifier of the first identifier-type and save the result as a first URI-specific core price result;
- receive a third and fourth set of price and product attributes for a second product, which attributes comprise: a third identifier of the first identifier-type, a fourth identifier of the second identifier-type, and a second category in the category taxonomy;
- perform a second URI-specific price analysis of price values in the third and fourth sets of price and product attributes to identify changes in price for the second product and associate the result with the third identifier of the first identifier-type and save the result as a second URI-specific core price result;
- when the second identifier and the fourth identifiers are the same, perform a first non-URI-specific price analysis utilizing the first and second URI-specific core price results to identify changes in price according to the second identifier-type and save the result as a first non-URI-specific core price result;
- save and index the output of the URI-specific and non-URI-specific price analyses in a first file structure and make the first file structure available to be searched substantially as the sets of product and price attributes are received;
- perform a meta-analysis utilizing the URI-specific and non-URI-specific core price results to identify what price and product attributes across the datasets are associated with the changes in price; and
24. A non-transient computer-readable storage medium having stored thereon instructions that, when executed by a processor, configure the processor to: save and index the output of the meta-analysis as a second file structure and make the second file structure available to be searched.
- receive a first and a second set of price and product attributes for a first product, which attributes comprise: a first identifier of a first identifier-type derived from a URI which links to a webpage offering the product for sale, a second identifier of a second identifier-type assigned to all instances of the product as offered for sale at any URI, and a first category in a category taxonomy;
- perform a first URI-specific price analysis of price values in the first and second sets of price and product attributes to identify changes in price for the first product and associating the result with the first identifier of the first identifier-type and save the result as a first URI-specific core price result;
- receive a third and fourth set of price and product attributes for a second product, which attributes comprise: a third identifier of the first identifier-type, a fourth identifier of the second identifier-type, and a second category in the category taxonomy;
- perform a second URI-specific price analysis of price values in the third and fourth sets of price and product attributes to identify changes in price for the second product and associate the result with the third identifier of the first identifier-type and save the result as a second URI-specific core price result;
- when the second identifier and the fourth identifiers are the same, perform a first non-URI-specific price analysis utilizing the first and second URI-specific core price results to identify changes in price according to the second identifier-type and save the result as a first non-URI-specific core price result;
- save and index the output of the URI-specific and non-URI-specific price analyses in a first file structure and make the first file structure available to be searched substantially as the sets of product and price attributes are received;
- perform a meta-analysis utilizing the URI-specific and non-URI-specific core price results to identify what price and product attributes across the datasets are associated with the changes in price; and
Type: Application
Filed: Jul 25, 2013
Publication Date: Jan 30, 2014
Inventors: Satyanarayana Rao KALIKIVAYI (Chennai), Mohammed J. ZAHOOR (Chennai), Sanjay PARTHASARATHY (Bellevue, WA)
Application Number: 13/951,248
International Classification: G06Q 30/02 (20060101);