SYSTEM, METHOD, AND APPARATUS FOR DATA COLLECTION AND AGGREGATION

Abstract

This disclosure relates to a system, apparatus, or method for satisfying a pet healthcare need. Certain non-limiting modalities include collecting data from one or more disparate data sources and aggregating the collected data from the one or more disparate sources into a data lake. Certain non-limiting modalities also include removing personally identifiable information from the aggregated data in the data lake and classifying the removed data into one or more datasets based on one or more attributes of the pet or a pet owner. In addition, such non-limiting modalities may include creating a profile of a pet based on the data included in the one or more datasets and determining a pet product to help satisfy the pet healthcare need.Additionally, certain non-limiting modalities may include targeting the pet owner with an advertisement for the pet product.

Description

SYSTEM, METHOD, AND APPARATUS FOR COLLECTION AND AGGREGATION RQfrC ίΠ / ΖΖηΖ / Ε / ΥΙΛΙ DATA Cross-reference to related applications [1] This application claims priority for United States of America Provisional Application No. 63 / 038,568, filed on June 12, 2020, the contents of which are hereby incorporated by reference in their entirety. FIELD OF INVENTION [2] This disclosure relates to the aggregation and linking of data from one or more sources, more particularly to the linking of pet care-related data from one or more sources. BACKGROUND [3] The digital data footprint created by pet owners in today's world is growing. Like any other consumer group, pet owners make numerous purchases online and in pet stores. These pet products are highly varied, ranging from food to toys and clothing. However, beyond traditional consumer goods, pet owners are clearly positioned to create digital data related to their pets. For example, some owners attach health or location monitoring devices to their pets. In addition, some owners use services for DNA or genetic testing of their pets. These devices and services, as well as other available services used by pets, can offer unique information about a particular pet, far beyond what can be estimated from traditional consumer data.Given the large volume of data being collected, it has become increasingly difficult to refine this data for the benefit of both the pet and the pet owner. [4] Consequently, there is a persistent demand in the pet care industry for a device, system, and method for analyzing pet care-related data. In particular, there remains a need to aggregate and link a diverse set of digital data related to pet care. Doing so can help owners better understand their pets' needs or preferences, leading pet owners to better care for their pets' health and well-being. BRIEF DESCRIPTION OF THE INVENTION [5] This disclosure relates to a system, apparatus, or method for aggregating and linking a diverse set of data related to pet care, resulting in a better understanding of the needs and desires of a particular pet. Certain non-limiting modalities may include a method for meeting a pet's health care need. RQfrC ίη / ZZΖΠZ / E / YΙΛΙ The limitations of the method may include collecting data from one or more disparate data sources and aggregating the collected data from the one or more disparate sources into a data lake. The method may also include removing personally identifiable information from the aggregated data in the data lake and classifying the removed data into one or more datasets based on one or more attributes of the pet or a pet owner. Furthermore, the method may include creating a profile of a pet based on the data included in the one or more datasets and determining a pet product to help meet the pet's healthcare needs. Additionally, the method may include targeting the pet owner with an advertisement for the pet product. [6] In certain non-limiting embodiments, the method may additionally include displaying the pet product advertisement to the pet owner through a graphical user interface on a terminal device. The method may also include reformatting data from one or more disparate data sources to a uniform level of quality. Furthermore, the method may include predicting a target audience for the pet product advertisement, where the target audience may include the pet owner and one or more other pet owners. In some embodiments, the method does not RQfrC ίη / ZZΖΠZ / E / YΙΛΙ Limiting, the data may include at least one of the following: first-party, second-party, or third-party data sources. The one or more disparate data sources may include a health and location monitoring device, a DNA or genetic testing service, or a point-of-sale device in a veterinary hospital. The collected data may include pet health information. In other non-limiting modalities, aggregation, deletion, classification, creation, and determination may be performed by a consumer data platform or a data management platform. Determination or prediction may use a machine learning algorithm for target audience prediction or pet product determination to help meet pet healthcare needs.The one or more data sets may include at least one of the pet's health, pet ownership, physical traits, or pet behavior. [7] In certain non-limiting embodiments, a system for satisfying a pet health care need may include at least a memory comprising computer program code, and at least one processor. The computer program code may be configured, when executed by at least one processor, to cause the apparatus to collect data from one or more disparate data sources, and aggregate the data collected from the one or more disparate sources. RQfrC iP / ZZΖ / E / YILI in a data lake. The computer program code can also be configured, when executed by at least one processor, to cause the device to erase personally identifiable information from the data aggregated in the data lake, and to classify the erased data into one or more datasets based on one or more attributes of the pet or a pet owner. In addition, the computer program code can also be configured, when executed by at least one processor, to cause the device to create a profile of a pet based on the data included in the one or more datasets, and to determine a pet product to help meet the pet's healthcare needs.Additionally, the computer program code can also be configured, when run by at least one processor, to cause the device to address the pet owner with an advertisement for the pet product. [8] According to certain modalities, a non-transient, computer-readable medium encodes instructions that, when executed on hardware, perform a process to satisfy a pet healthcare need. The process may include collecting data from one or more disparate data sources and aggregating the collected data from the one or more disparate sources into a data lake. The process may also include deleting personally identifiable information from the aggregated data in the RQfrC iP / ZZΖ / E / YILI data lake, and classify the deleted data into one or more datasets based on one or more attributes of the pet or a pet owner. Furthermore, the process may include creating a pet profile based on the data included in the one or more datasets, and determining a pet product to help meet the pet's healthcare needs. Additionally, the process may include targeting the pet owner with an advertisement for the pet product. [9] An apparatus, in certain modalities, may include a computer program product that encodes instructions for processing data from a pet product tested according to a method to satisfy a pet health care need. Certain non-limiting modalities of the method may include collecting data from one or more disparate data sources and aggregating the collected data from the one or more disparate sources into a data lake. The method may also include deleting personally identifiable information from the data aggregated in the data lake and classifying the deleted data into one or more datasets based on one or more attributes of the pet or a pet owner. In addition, the method may include creating a profile of a pet based on the data included in the one or more datasets and determining a product RQfrC ίη / ZZΖΠZ / E / YΙΛΙ for pets to help meet the pet's healthcare needs. In some forms, the method may include identifying a pet product to help meet any pet service or pet product needs. Additionally, the method may include targeting the pet owner with an advertisement for the pet product.

[10] It should be understood that both the above general description and the following detailed description are exemplary and are intended to provide further explanation of the disclosed subject matter claimed. BRIEF DESCRIPTION OF THE DRAWINGS

[11] FIGURE 1 illustrates a system or apparatus for the collection or aggregation of data according to certain non-limiting modalities;

[12] FIGURE 2 illustrates a system for the collection or aggregation of data according to certain non-limiting modalities;

[13] FIGURE 3 illustrates a system for the collection or aggregation of data according to certain non-limiting modalities;

[14] FIGURE 4 illustrates a linking system according to certain non-limiting modalities;

[15] FIGURE 5 illustrates a system or method for aggregation and linking according to certain non-limiting modalities; RQfrC ίη / ΖΖΠΖ / Ε / ΥΙΛΙ

[16] FIGURE 6A illustrates a system or method for aggregation and linking according to certain non-limiting modalities;

[17] FIGURE 6B illustrates a system or method for aggregation and linking according to certain non-limiting modalities;

[18] FIGURE 7 illustrates a system or method for aggregation and linking according to certain non-limiting modalities.

[19] FIGURE 8 illustrates a diagram according to certain non-limiting modalities;

[20] FIGURE 9 illustrates a flowchart according to certain non-limiting modalities;

[21] FIGURE 10 illustrates a diagram of a system according to certain non-limiting modalities; and

[22] FIGURE 11 illustrates a diagram of an apparatus or system according to certain non-limiting modalities. DETAILED DESCRIPTION

[23] There remains a need to aggregate and link a diverse or disparate set of data related to pet care, resulting in a better understanding of the needs and desires of a given pet. Specifically, there remains a need for techniques that allow the aggregation and linking of a diverse set of digital data in order to predict or determine products RQfrC ίΠ / ZZΖ / E / YΙΛΙ for pets related to the health care needs of a pet. The subject matter currently disclosed addresses these and other needs.

[24] For clarity and not as a limitation, this detailed description is divided into the following sub-portions: A. Definitions; B. Data collection or aggregation; C. Linking data and predicting audiences or products; D. Results of the system or method described herein; E. Flowchart of the method described herein; and F. Diagrams of the system and apparatus described herein.

[25] A. Definitions

[26] The terms used in this specification generally have their ordinary meanings in the art, within the context of this disclosure and in the specific context where each term is used. Certain terms are discussed below, or elsewhere in the specification, to provide additional guidance in describing the compositions and methods of the disclosure and how to make and use them.

[27] As used in the specification and accompanying claims, the singular forms a, an, the and the include plural referents unless the context clearly indicates otherwise. Thus, for example, RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ the reference to a compound includes mixtures of compounds.

[28] The terms animal or pet, as used in accordance with this disclosure, refer to domestic animals, including, but not limited to, domestic dogs, domestic cats, horses, cows, ferrets, rabbits, pigs, rats, mice, gerbils, hamsters, goats, and the like. Domestic dogs and cats are particular, non-limiting examples of pets. The term animal or pet, as used in accordance with this disclosure, may additionally refer to wild animals, including, but not limited to, bison, elk, deer, ducks, poultry, fish, and the like.

[29] As used herein, the terms comprise, understand, or any other variation thereof, are intended to cover a non-exclusive inclusion, so that a process, method, article, system, or apparatus comprising a list of items does not include only those items, but may include other items not expressly listed or inherent in such process, method, article, or apparatus.

[30] In the description detailed here, references to modality, a modality, a modality, in several modalities, certain modalities, some modalities, other modalities, certain other modalities, etc., indicate that the modality(ies) described may include a particular feature, structure, or characteristic, but each modality RQfrC iΠ / ZZΖ / E / YΙΛΙ might not necessarily include the particular feature, structure, or characteristic. Furthermore, such phrases do not necessarily refer to the same modality. Additionally, when a particular feature, structure, or characteristic is described in relation to one modality, it is stated that it is within the knowledge of a person skilled in the art to affect such feature, structure, or characteristic in relation to other modalities, whether explicitly described or not. After reading the description, it will be evident to a person skilled in the relevant art(s) how to implement the disclosure in alternative modalities.

[31] In general, terminology can be understood, at least in part, from its use in context. For example, terms such as and, or, or, as used here, can encompass a variety of meanings that may depend, at least in part, on the context in which such terms are used. Typically, or, if used to associate a list, such as A, B, or C, is intended to mean A, B, and C, used here in the inclusive sense, as well as A, B, or C, used here in the exclusive sense. Furthermore, the term one or more, as used here, depending, at least in part, on the context, can be used to describe any one trait, structure, or characteristic in a singular sense or can be used to describe combinations of traits, structures, or characteristics in a plural sense. Similarly, the RQfrC Ln / Zznz / E / YIAI Terms such as a, an, the, or a, again, can be understood to convey a singular or plural usage, depending, at least in part, on the context. Furthermore, the term based on can be understood as not necessarily intended to convey an exclusive set of factors and, instead, may allow for the existence of additional factors not necessarily expressly described, again, depending, at least in part, on the context.

[32] The term "data" includes, for example, without limitation, any type of digital or non-digital data relating to a pet or a pet product. In certain modalities, data may be any measurement relating to the pet's health, breed, physical attributes, disease, diagnosis, location, activity, nutritional intake, prescriptions, wellness plan, genetics, and / or deoxyribonucleic acid (DNA), or any other information relating to the pet. In some other modalities, data may include any information relating to the pet owner's habits, purchases, or activity, whether or not related to the pet itself. Data may also include searches, browsing, or purchasing habits on one or more websites.

[33] The term first-party data may include data directly collected and / or held by a company. RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ of the first, for example, can be collected from consumers who use products or services offered by a particular company.

[34] The term third-party data may include data collected by other companies. Third-party data can be accessed or purchased using a data management platform. For example, a company may purchase third-party data to expand its target audiences.

[35] The term pet product includes, for example, without limitation, any type of product, service, or equipment that is designed, manufactured, or used by a pet. For example, a pet product may be a toy, a chew toy, food, clothing, a collar, medication, and / or a health or location monitoring device. In another example, a pet product may include a pet DNA or genetic testing service.

[36] The term pet owner may include any person, organization, or group of people responsible for the care of a pet.

[37] The term digital-side platform can refer to a system or software that facilitates the purchase of digital advertising. In other words, a digital-side platform allows advertisers to purchase digital advertising, also known as impressions, across a variety of websites. Advertising can be targeted to specific users based on RQfrC iP / ZZΖ / E / YILI data or information related to users. For example, advertising can be specifically targeted to a pet owner based on data related to the pet. In certain modalities, the digital-side platform can help manage an inventory of available advertising space provided by publishers.

[38] The term real-time bidding can refer to a scheduled auction that allows buyers to purchase available advertising space. In certain modalities, real-time bidding between one or more buyers can be managed using a digital platform.

[39] The term data management platform can refer to a system or software used to collect, organize, manage, and / or activate large datasets from disparate sources. A data management platform, for example, may include one or more processes for collecting, unifying, organizing, activating, and / or analyzing data. For instance, data management platforms may utilize one or more artificial intelligence or machine learning algorithms to process and analyze datasets. In some forms, data management platforms collect anonymous digital and / or web data. Data management platform records can be updated discreetly using a batch process.

[40] The term customer data platform can be a system or software used to collect data and create RQfrC iP / ZZΖ / E / YILI is a unified customer database accessible to other systems. In certain configurations, a customer data platform can collect raw data from one or more sources, clean and / or reformat the data, and aggregate or combine the data to form a single customer or pet profile. Consequently, customer data platforms can collect data that is linked to an identifiable person, consumer, or pet. In some configurations, the customer data platform can collect data continuously or discreetly to keep profiles up-to-date and accurate.

[41] The term data lake can refer to a database, a data library, a data repository, or any other storage medium. In certain configurations, the data lake may be centralized or decentralized. The data maintained within the data lake may be structured, unstructured, or semi-structured. For example, the data lake may include raw first-party data, which can be considered unstructured.

[42] The term personal data can be any information that can be used to infer or determine the identity of a person or pet. In certain contexts, personal data may be referred to as personally identifiable information or information that can be personally identified. For example, a name, identification number Governmental RQfrC ίη / ZZΖΠZ / E / YΙΛΙ, date and place of birth, mother's maiden name, and / or biometric records may be considered personal data. Other personal information may include an Internet protocol address, geolocation, and / or a patient ID provided by a healthcare facility. Any other information that can be linked to a person or pet, such as medical, educational, financial, and / or employment information, may also be considered personal information.

[43] The term geofencing zone, geofencing location, or geofencing area can refer to a predetermined, preselected, or known geographic location or area selected by the user. The geofencing location or zone can be used to define a familiar geographic location or area for the pet or animal, such as the residence of the pet owner or caregiver. In some examples, the geofencing location or zone can be determined using a Global Positioning System (GPS) or a Global Navigation Satellite System (GLONASS) receiver to determine the latitude and longitude of a device. In other examples, the presence, location, or geofencing zone can be determined using a predetermined, preselected, or known Service Set Identifier (SSID) of a Wireless Land Area Network (WLAN). When the geofencing location or zone RQfrC iP / ZZΖ / E / YILI is determined using an SSID scan, or any other feature or measurement associated with the WLAN; the location or geofencing zone can be referred to as a beacon zone. In other words, in certain non-limiting modes, a client station can be determined to be in a given zone or location when it is within the transmission range of a given access point. To determine whether the client station is located inside or outside a geofencing location or zone, the client station can transmit probe requests and receive probe responses from an access point (AP) on the WLAN network.

[44] The term terminal device refers to, for example, without limitation, a personal computer, laptop, workstation, television, mobile device, terminal device, or any other user equipment. In some non-limiting examples, the terminal device may include a graphical user interface used to display an advertisement related to a pet product. In certain modalities, a web-based application may be installed on the user's terminal device. In other non-limiting modalities, the terminal device may be a pet monitoring device with an accelerometer or microelectromechanical system (MEMS) device.

[45] B. Data collection or aggregation

[46] In certain non-limiting modalities, the data are RQfrC iP / ZZΖ / E / YILI can collect data from one or more sources. The collected data can then be aggregated and linked to create a user, client, or pet profile. The data may include first-party data, second-party data, and / or third-party data. For example, first-party data may be collected from a health and location monitoring device, a DNA or genetic testing service, and / or a point-of-sale (POS) device in a veterinary hospital.

[47] FIGURE 1 illustrates a system or apparatus for collecting or aggregating data according to certain non-limiting modalities. In particular, FIGURE 1 illustrates a pet health and / or location monitoring device 110, which can be a source from which data can be collected. The pet health and / or location monitoring device 110 can be a computing device used to monitor and / or track pet activity. In some non-limiting modalities, the pet health and / or location monitoring device 110 can be configured to collect data generated by various hardware or software components.Examples of hardware or software components may include a global positioning system (GPS) receiver and / or one or more detectors or transducers, such as an accelerometer, a gyroscope, or any other device or component used to record, collect, or receive data relating to the. RQfrC iΠ / ZZΖηZ / E / YILI movement or activity of the pet that wears, or otherwise carries, the monitoring device 110. The pet health and / or location monitoring device 110 may include one or more processors, memory, transducer, and / or any other hardware component used to receive, collect, process, store, and / or transmit data.

[48] ​​Examples of the 110 pet health and / or location monitoring device are described in U.S. Patent Application No. 16 / 570,771, U.S. Patent Application No. 16 / 182,057, now U.S. Patent No. 10,708,709, U.S. Patent Application No. 15 / 291,882, now U.S. Patent No. 10,142,773, U.S. Patent Application No. 15 / 287,544, U.S. Patent Application No. 14 / 231,615, now U.S. Patent No. 10,420,401, and U.S. Provisional Applications No. 62 / 970,575, 62 / 867,226, and 62 / 768,414, United States of America Design Application No. 29 / 696,311, United States of America Design Application No. 29 / 696,315 and International Patent Applications No. PCT / US17 / 55220 and PCT / US17 / 55224.The applications and patents mentioned above are incorporated herein by reference in their entirety. RQfrC ίη / ΖΖΠΖ / Ε / ΥΙΛΙ

[49] In certain, non-limiting modalities, the 110 health and location monitoring device may initially collect information from the 101 retail channel and / or 102 consumer data. The 101 retail channel information may be any information or data related to the merchant or retailer from whom the 110 health and location monitoring device was purchased. The 101 retail channel may, for example, include one or more wholesalers, club stores, drugstores, health product stores or pharmacies, specialty pet stores, e-commerce, or pet-related stores. The 110 health and location monitoring device may be purchased online, through a retailer or merchant, or through any other means.Consumer data (102) can be any information, personal or otherwise, related to either the pet owner who purchases the monitoring device (110) or the pet wearing the monitoring device (110). Consumer data (102) can be determined from the pet owner during the initial activation or setup of the monitoring device (110).

[50] Consumer data, for example, may include household income, home ownership, spending habits, geolocation, lifestyle attributes, weather patterns or information, data Environmental factors and household purchasing habits are also relevant. Other consumer data may include demographics, lifestyle, vehicle data, census data, creditworthiness, socioeconomic data, consumer attitudes, behaviors, preferences, purchasing habits, lifestyle, health, well-being, finances, and / or wealth. In certain non-limiting modalities, consumer data may include loyalty card data, which many retailers use to identify an individual regardless of the offer. Loyalty card data may include details related to items purchased. In some non-limiting modalities, consumer data may include the transactional total from payment card issuers and / or panel data from consumer panels that scan all purchases at retail locations. Panel data may be used to project sales.In other, non-limiting, forms, consumer data 102 may include exposure data, print circulation data for newspapers and magazines, mobile and radio data, and / or Nielsen media data. Exposure data may include publisher data, such as page views, stores visited, and length of time spent viewing a particular webpage or website. Nielsen media data may include television viewer panel data. In some other forms, this does not apply. RQfrC ίΠ / ZZηZ / E / YΙΛΙ Limitations, consumer data 102 may include purchase data, including any data related to the consumer's purchases. For example, purchase data may include card data from frequent shoppers, such as UPC scans or card IDs, for which promotions and messages can be activated based on the shopper's purchase information. Promotions and / or messages may be delivered at the point of sale, for example, by means of a high-speed, full-color printer.

[51] The 110 health and location monitoring device may collect network connection or geographic location information. Network connection information may be used to determine the location of the 110 health and / or location monitoring device, as well as other network metrics that may provide information related to the pet or pet owner. In certain, non-limiting, modes, the 110 health and location monitoring device may collect information such as the 121 Global System for Mobile Communications (GSM) frequency, 122 GPS tracking, WLAN, 123 fourth-generation (4G) or fifth-generation (5G) connection, 124 geofencing, 127 park location, 128 retail store location, and 129 lodging location.GSM is a standard developed by the European Telecommunications Standards Institute (ETSI) to define the protocols for second-generation (2G) digital cellular networks. Location 128. The retail store location RQfrC ίη / ZZΖΠZ / E / YΙΛΙ may include a location and / or identity of the retail store visited by a pet or the pet owner. The boarding location 129 may include a location and / or identity of a pet boarding facility. The park location 127 may include an outdoor or indoor park visited by a pet wearing the health and location monitoring device 110.

[52] Other information collected by the health and location monitoring device 110 may include pet activity 125. Pet activity 125 may include any body movement detectable by the health and location monitoring device 110. For example, pet activity 125 may include scratching, licking, pacing, drinking, eating, sleeping, and shaking, and / or any other body movement associated with an action performed by the pet. The health and location monitoring device 110 may also be used to identify pet activity 125, such as a pet jumping, moving excitedly in search of food, eating voraciously, and / or drinking from the bowl on the wall. Pet activity 125, also referred to as pet behavior, may be species-specific.For example, there may be behaviors attributable to cats and other behaviors attributable to dogs. While one or more behaviors... RQfrC ίΠ / ZZΖ / E / YΙΛΙ may overlap; some behaviors may be exclusive to cats or dogs. One or more behaviors, in certain modalities, may be a symptom, or indicative of a symptom, associated with a health condition or disease. In some non-limiting modalities, pet scratching 126 may be collected separately from other pet activities.

[53] The raw data collected from the health and location monitoring device 110 can be forwarded to the data lake 130. The data can then be forwarded to the data consent framework 140. The data consent framework 140 can help ensure that data handling complies with regulatory guidelines. After the data is collected and cleaned, its accuracy and reliability can be ensured. For example, the data consent framework 140 could be a framework created by the Internet Advertising Bureau (IAB) that uses a transparency and consent chain (TC chain). The TC chain can be used to encapsulate relevant details on how transparency and consent can be established and coded as they apply to purposes, special purposes, features, and special characteristics.In particular, the TC string can encapsulate and encode all information disclosed to a user and / or the expression of their preferences for the processing of their personal data under. RQfrC iP / ZZΖ / E / YILI the General Data Protection Regulation (GDPR) or other international, national, state, or local regulations. Using a Consent Management Platform (CMP), the information can be captured in a compact, encoded HTTP transferable string. This string can enable the communication of transparency and consent information to one or more entities or vendors that process a user's personal data, including, for example, the California Consumer Privacy Act. The entities or vendors can then decode a TC string to determine whether they can process a user's personal data for their purposes based on existing international, national, state, or local regulations. The concise string data format allows a CMP to persist and retrieve user preferences, as well as transfer that information to one or more entities or vendors.

[54] In certain non-limiting modalities, the TC string may include one or more of the following: general metadata, user consent, legitimate interest, published restrictions, published transparency and consent, out-of-band legal bases, and / or jurisdiction-specific disclosures. General metadata may include markers that indicate details about a TC string, such as the encoding version, the last update, when it was initially created, details about the conditions of the values ​​of RQfrC Ln / Zznz / E / YIAI transparency and consent, for example, the global list of vendors, and / or the CMP used. Consent for use may be the user's expression of consent given for the processing of their personal data, whether by purpose or by vendor. Legitimate interest, for example, may be the record of a CMP that has established transparency of legitimate interest for a vendor or purpose, and / or whether the user exercised their right to object. Publisher transparency and consent may be a segment of a TC string that publishers may use to establish transparency with or obtain user consent for their own legal bases to process personal data and / or share personal data with vendors. The out-of-band legal basis may be at least two segments used to express that a vendor is using out-of-band legal bases to process personal data.For example, the first segment may be a list of vendors disclosed to the user, while the second may be a list of vendors that the publisher may allow to use legal out-of-band bases.

[55] Once processed through the 140 data consent framework, the information can be forwarded to the 530 prediction tool, as shown in FIGURE 5. The 530 prediction tool, which may be a customer data platform (CDP) and / or may use intelligence RQfrC ίΠ / ZZηZ / E / YILI artificial intelligence or machine learning can link data and create one or more consumer or user profiles. Data linking can utilize an identity resolution process. The prediction tool 530, for example, can test the predictability of dog owners who have visited a veterinarian in the past year, the predictability of dog owners with one or more dogs who have purchased skin treatment products in the past year and / or visited a veterinarian in the past year. In yet another example, the prediction tool 530 can test the predictability of dog owners with one or more dogs who are in the toy dog ​​category, purchased skin treatment products in the past year, and / or visited a veterinarian in the past six months, or any other time period.Different combinations of data points can be used to predict which provides the best, largest, or most effective response to a consumer advertisement and / or, ultimately, the impact on sales.

[56] In some non-limiting modalities, the 530 prediction tool may use analysis of covariance (ANCOVA), which can be a general linear model using analysis of variance (ANOVA) and / or regression. ANCOVA can be used to assess whether the means of a dependent variable (DV) are equal across the levels of a categorical independent variable (IV), which may be referred to as a RQfrC ίΠ / ZZΖηZ / E / YILI factor or treatment, while statistically controlling for the effects of other continuous variables that are not of primary interest, which may be referred to as covariates (CV) or nuisance variables. Although ANCOVA can be used when there are differences between your reference groups, it can also be used in pre-test or post-test analysis when regression to the mean may affect your post-test measurement.

[57] In certain non-limiting modalities, ANCOVA can be used as an extension of multiple regression analysis. Similar to regression analysis, ANCOVA can be used to assess how independent variables affect dependent variables. In non-limiting modalities, ANCOVA can eliminate any effects of covariates. For example, ANCOVA can be used to determine whether a new drug works for depression, with three treatment groups and one control group. ANCOVA can be used to determine whether the treatment works while controlling for other factors that may influence the outcome, such as family life, employment status, or drug use. In other non-limiting modalities, ANCOVA can be used as an extension of ANOVA to control for covariates and / or test combinations of categorical or continuous variables as predictors.In some non-limiting modalities, covariates may be variables of interest, in. RQfrC ίΠ / ΖΖηΖ / Ε / ΥΙΛΙ instead of variables that are controlled.

[58] ANCOVA can be used to test for or explain within-group variance. For example, ANCOVA can take the unexplained variances from the ANOVA test and explain those unexplained variances with confounding variables, or one or more distinct covariates. Weak covariates can reduce the statistical power of the test, while strong covariates can increase it. In certain non-limiting modalities, the following process or method can be performed: perform a regression between the independent and dependent variables, identify the residuals from the regression results, run an ANOVA on the residuals, and incorporate the assumptions for the ANCOVA.Assumptions may include at least two independent variables as categorical variables, dependent variables and covariates may be continuous, as measured on an interval scale or a ratio scale, and / or ensure that observations are independent.

[59] After running the ANVOCA analysis, a p-value can be calculated. In certain modalities, the p-value can represent the probability of obtaining results at least as significant as the observed results of a statistical hypothesis test, where the null hypothesis can be assumed to be correct. In certain non-limiting modalities, the p-value The p-value can be used as an alternative to rejection points to provide the smallest significance level at which the null hypothesis can be rejected. The p-value can be inversely proportional to the likelihood that the alternative hypothesis is correct, meaning that the smaller the p-value, the greater the probability that the alternative hypothesis is correct. In certain non-limiting modalities, a calculated p-value indicating approximately 80% confidence can be selected. In alternative modalities, a calculated p-value indicating greater than approximately 80% confidence can be selected. In such non-limiting modalities, the p-value calculation can occur multiple times after the ANCOVA analysis. In other non-limiting modalities, the calculated p-value indicates that a confidence level of less than approximately 80% can be selected.In alternative modalities, a new ANCOVA analysis may be performed after calculating a p-value that indicates less than a desired level of confidence. In other alternative modalities, calculating a p-value is not necessary at all.

[60] In certain non-limiting modalities, data cleansing or debugging may be used. Data cleansing or debugging can be used to detect, correct, and / or remove corrupt or inaccurate records from a recordset, table, or database. For example, the Data cleansing or debugging can include identifying incomplete, incorrect, inaccurate, or irrelevant parts of the data and / or replacing, modifying, or deleting dirty or coarse data. Data cleansing can be performed interactively with data dispute tools or as batch processing through scripts. In certain non-limiting modes, data cleansing or debugging may not produce any changes to the data. For example, data cleansing or debugging might detect that there are no errors in the data. Alternatively, and without limitation, data cleansing or debugging can be configured to ignore certain errors. Additionally, data cleansing or debugging can be configured to identify commonly misidentified errors, for example, data that is often identified as an error but should still be included in the resulting dataset.

[61] After cleansing, a dataset may be consistent with other similar datasets in the system. The inconsistencies detected or eliminated could have originally been caused by, for example, a user input error, corruption in transmission or storage, or different data dictionary definitions of similar entities in different storage locations. In some non-limiting modalities, data cleansing or debugging may differ from data validation in that validation may mean that data is rejected from RQfrC iP / ZZΖ / E / YILI system at the input and can be performed at the time of entry, rather than in batches of data. Data cleansing, for example, can involve removing typographical errors or validating and / or correcting values ​​against a known list of entities. Validation can be strict, such as rejecting any address that may not have a valid postal code, or fuzzy, such as correcting records that only partially match existing or known records. In certain non-limiting modalities, data cleansing solutions can clean data by matching it against a validated dataset. Data cleansing can also include data enhancement, where the data can be made more complete by adding related information. For example, addresses can be supplemented with one or more telephone numbers.In some other approaches, data cleansing may include data harmonization (or normalization), which can involve combining data from different file formats, naming conventions, and columns, and / or transforming the data into a cohesive dataset. For example, abbreviations such as st, rd, etc., can be expanded to street, road, etc.

[62] An identity can refer to the anonymous profile of a consumer collected through one or more user devices or equipment. For example, constructing a person's identity may include combining data collected from RQfrC ίη / ZZΖΠZ / E / YΙΛΙ laptop browsers, cell phones, email subscriptions, and / or offline purchases are pooled together. The combined data can then be anonymized so that marketers, or any other entity or person with access to the identity, would not have access to the consumer's personally identifiable information. Personally identifiable information, for example, might be a name, email address, phone number, or mailing address. In some, but not limited to, other consumer-related information, such as gender, age, location, and / or the consumer's browsing or purchasing activities, may be included as part of the identity, which can be reviewed by marketers or any other entity or person with access to the identity.In certain non-limiting ways, identity resolution can be the compilation of all information into a single profile. For example, identity resolution might indicate that a consumer conducts business activities in one browser but personal activities, such as personal purchases, in a different browser. In yet another example, identity resolution might indicate that the consumer's mobile device contains more data or better reflects the consumer's off-hours habits, such as preferred television programs or items. RQfrC ίΠ / ZZΖ / E / YILI personal information that is searched across one or more internet browsers. Identity resolution provides marketers, or any other entity or person with access to identity resolution, with a holistic view of the consumer. Identity resolution can therefore connect different identifiers across multiple platforms and / or devices to enable people-based targeting, personalization, or measurement. In some, but not limited to, modalities, identity resolution can be built or aggregated in real time.

[63] FIGURE 2 illustrates a system for collecting or aggregating data according to certain non-limiting modalities. In particular, FIGURE 2 illustrates a 210 service for pet DNA or genetic testing. The 210 service for DNA or genetic testing may be one of the sources from which data can be collected and aggregated. Under certain non-limiting modalities, the 210 service for DNA or genetic testing may initially collect information from the 201 retail channel and / or 202 consumer data. The 201 retail channel and 202 consumer data may be similar to the 101 retail channel and 102 consumer data shown in FIGURE 1.

[64] The 210 DNA or genetic testing service may be in the form of an at-home testing kit provided to pet owners and / or a test performed by a veterinarian. For those modalities that involve a RQfrC iP / ZZΖ / E / YILI Home Test Kit: A pet owner can purchase a test kit. The pet owner can then use one or more swabs from the test kit to collect skin cells from the pet's mouth. For example, the pet owner can swab the inside of the pet's cheek. Once the sample is collected, the pet owner can mail it to a laboratory for analysis. In certain non-limiting modalities, a pet owner may be required to use the internet to activate the home test kit. As part of the activation, the pet owner may be asked to answer one or more questions related to the pet or the pet owner. These answers may be collected as consumer data.

[65] DNA or genetic testing services 210 can be used to determine and collect a variety of biological, physical, health, or ancestry-related information about the pet. For example, the pet's breed 221, potential diseases 223, health conditions 224, DNA information 225, physical traits 226, the pet's relatives 227, and / or pedigree 229 can be collected. Other information may include direct-to-consumer sales 222, paid offerings 228, and / or testing kits 231. For example, a direct-to-consumer sale 222 may include any product or service sold on an online website or in a RQfrC iP / ZZΖ / E / YILI store. The paid offer 228 may include information collected from the purchase of a product or service using a payment card. The test kit 231 may be a DNA or disease test kit. The raw data collected from the DNA or genetic testing service 210 may be forwarded to the data lake 240. The data may then be forwarded to the data consent framework 250, which may be similar to the data consent framework 140. Once processed through the data consent framework 250, the information may be forwarded to the prediction tool 530, as shown in FIGURE 5. The prediction tool 530, which may be a CDP, may then link the data and create one or more consumer or user profiles.

[66] The breed 221 of the pet may include breeds of dogs, breeds of cats, or a breed for any other pet or animal. Some examples of dog breeds may include Affenpinscher, Afghan Hound, Kuchi, Aidi, Airedale Terrier, Akbash, Akita, Spanish Alano, Alaskan Husky. Some examples of cat breeds may include Abyssinian, Aegean, American Bobtail, American Curl, American Shorthair, American Roughhair, Aphrodite Giant, Arabian Mau, and Asian cat. Potential diseases 223 may be diseases specific to dogs, cats, or any other pet. For example, diseases of dogs may include abnormal elbow development, abnormal eyelids, RQfrC ίη / ZZΖΠZ / E / YΙΛΙ abnormal growth in the lower intestines, abnormal heart rate, abnormal molar development, abnormal passage between the artery and vein, abnormal passage between the mouth and nasal cavity, abnormal protein production, abnormal urine output due to urinary bladder dysfunction, or any other related disease in dogs. Disease in cats, for example, may include abnormal cavity swelling, abnormal diaphragmatic opening, abnormal eyelid, abnormal growth in the lower intestines, abnormal heart rate, abnormal passage between the artery and vein, abnormal passage between the mouth and nasal cavity in cats, abnormal protein production, abscesses, or any other related disease in cats.

[67] FIGURE 3 illustrates a system for collecting or aggregating data according to certain non-limiting modalities. In particular, FIGURE 3 illustrates information obtained from a pet healthcare facility, such as a veterinary clinic or hospital. The information can be obtained from the veterinary hospital's POS device, or any other device in the veterinary hospital with access to pet data, such as pet records or medical charts. The data collected or aggregated by the veterinary hospital's POS device can be a variety of biological, physical, or health-related data concerning the pet. For example, RQfrC ίη / ZZΖΠZ / E / YΙΛΙ attributes 321 of the pet, multiple pets 322, such as the number of pets owned by a given pet owner, prescription 323 of the pet, such as any medication, drug, or treatment prescribed to the pet, medical or health diagnosis 324 of the pet, records 325 of the pet's health, nutritional information 326 of the pet, such as any information related to the pet's nutritional intake, attachment 328 of the pet, wellness plans 331, and / or laboratory diagnoses 332. Attachment 328 of the pet may be the pet's adherence to a medical treatment or routine prescribed by the veterinarian and / or behavioral attachment to the pet owner.Other data collected may include household information, pet purchases made using the POS device, responses to advertisements displayed on the POS device or at the veterinary hospital, and the payment method, whether credit, debit, store credit, virtual currency, online payment, cash, and / or co-payment for pet owners with pet insurance. Pet purchases may include, for example, pet treats or pet food, toys, nutrition, snacks, clothing, collars, leashes, harnesses, cages, kennels, apparel, pharmaceuticals, beds, flea and tick control products, gates, doors, health and wellness products, bowls, and Internet of Things devices. RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ genetic testing products, insurance, wellness plans, and / or monitoring devices.

[68] 330 responses to advertisements may include, for example, responses to a post-campaign analysis (PCA). In certain, but not limited to, ways, a marketing or advertising campaign may use responses to advertisements to evaluate campaign performance in building brand awareness, influencing consideration, driving action, and / or engagement.The metrics used to evaluate the effectiveness of an advertising or marketing campaign can be, for example, cost per value view (CPW), the percentage of the target audience reached, frequency, which can be the time the ad is served to a consumer through one or more channels, the view completion rate, which can be a measurement of the percentage of the video viewed, positive earned media, which can be a measurement of where people have shared the content positively, and / or brand buzz, which can relate to the amount of conversation taking place around the brand involved in the campaign.In some non-limiting modalities, the metrics used to assess influence on consideration and / or action impulse may include, for example, quality score, click-through rate, impressions, average position, and / or conversion rate, which can be how many people do. RQfrC iP / ZZΖ / E / YILI Clicks on an advertisement lead to a desired action on the advertised website. In other, non-limiting modalities, the metrics used to evaluate action impulse and / or engagement may include, for example, total traffic, traffic by channel, bounce rate, which may be the percentage of visitors who leave before taking a desired action, conversions, which may be a quantifiable measure of how visitors have taken a desired action, or data capture, which may be the quality of the data obtained from visitors arriving at a website.

[69] A POS device may include an electronic cash register and / or software to coordinate the data collected from daily purchases. Retailers can enhance functionality by installing a network of data capture devices, such as card readers or barcode scanners. Consumers can be identified by the POS, for example, through the use of a customer or loyalty card, an email address, a telephone number, or payment card information. In some, but not limited to, POS devices can be used to monitor inventory and purchasing trends, help retailers avoid customer service problems such as out-of-stock sales, and tailor purchasing and marketing efforts. RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ to consumer behavior. Then, the collected data can be forwarded to the 340 data lake for aggregation. The 350 data consent framework can then receive or obtain the aggregated data from the 340 data lake.

[70] In certain non-limiting modalities, data collected from one or more of the sources shown in FIGURES 1, 2, and / or 3 may be healthcare data that includes personal or personally identifiable information (PII) of the pet or pet owner. For example, DNA 225 in FIGURE 2 and pet health records 325 in FIGURE 3 may include PII. This PII may be removed either before or after data aggregation in data lakes 130, 240, and 340. In certain modalities, data consent frameworks 140, 250, and / or 350 may be a data consent protocol (DCP) used to remove PII from aggregated or collected data. For example, consumer data may be moved to a restricted-access area with data / network security. Then, the names and / or addresses can be removed and a pseudo-anonymous identification can be assigned.Consumer data can then be used with pseudonymous identification. In some other modalities, the data consent framework 140, 250, and / or 350 can be a data management platform (DMP).

[71] For example, as described above, the framework RQfrC ίΠ / ΖΖηΖ / Ε / ΥΙΛΙ Sections 140, 250, and / or 350 of the data consent process can be used to remove PII from aggregated or collected data. In certain non-limiting modalities, PII can be transmitted to an area with restricted access, such as within a marketer's or entity's internal network, using network or data security. Names, addresses, or any other PII can be removed, and a pseudo-anonymous ID can be assigned. Any use of the consumer data can then be made using the pseudo-anonymous ID instead of any PII. The pseudo-anonymous ID can then be cross-checked with one or more third parties using, for example, the SHA-256 cryptographic hash algorithm.

[72] A cryptographic hash, which can also be referred to as a digest, can be a type of signature for a text or data file. The SHA-256 cryptographic hash algorithm, for example, can generate a nearly unique 256-bit (32-byte) signature for a text. A hash, in certain non-limiting forms, is not encryption, meaning that it cannot be decrypted back to the original text. Hashed versions of texts can be compared to each other, rather than comparing the original text.

[73] Any other data, whether first-party, second-party, and / or third-party data, may be collected and aggregated. Data may include, for example, animal symptoms, patient observations, health reports, and test results. RQfrC ίη / ZZΖΠZ / E / YΙΛΙ laboratory, treatment plans, nutritional records, activity-based information, animal genetics, interest-based reading by pet owners, media exposure by pet owners, pet owner purchasing habits, pet owner spending, and / or owner insurance contracts.

[74] Once collected and / or aggregated, the data in data lakes 130, 240, and 340 can be classified into datasets based on one or more attributes. Attributes can be categories or classifications that describe each dataset. For example, datasets can be attributed to pet health, pet owner (such as consumer data), the physical or behavioral traits of a dog or cat, and / or any other attribute. The health dataset might include information related to veterinary clinics and hospitals, such as consumer pet ownership data, diagnoses, number of visits in a given time period, prescriptions, pet body condition score, cat age range, dog age range, cat age category, and / or dog age category.A pet's body condition score can be considered the pet equivalent of the human body mass index. In certain non-limiting cases, a pet's body condition score can be determined by a professional. RQfrC Ln / Zznz / E / YIAI veterinarian to indicate a classification of thin, underweight, ideal, overweight, or obese. For example, the pet's body condition score may range from 1 to 5, with 1 (1 / 5) being very thin or thin, 2 (2 / 5) being underweight, 3 (3 / 5) being ideal weight, 4 (4 / 5) being overweight, and 5 (5 / 5) being obese. In another example, the pet's body condition score can range from 1 to 9, with 1 (1 / 9) being emaciated, 2 (2 / 9) being very thin, 3 (3 / 9) being thin, 4 (4 / 9) being underweight, 4.5 to 5 (4.5 or 5 / 9) being ideal weight, 6 (6 / 9) being overweight, 7 (7 / 9) being overweight, 8 (8 / 9) being obese, and 9 (9 / 9) being severely obese.

[75] In certain non-limiting modalities, a score of 1 / 5 or 1 / 9 may be assigned to pets with protruding ribs, spine, or pelvic bones, present muscle loss, and / or severe reduction of abdomen and dramatic waistline. A score of 1.5 / 5 or 2 / 9 may be assigned to pets with ribs, spine, or pelvic bones that are visible, but only minor muscle loss, and / or severe waistline and abdominal reduction. A score of 2 / 5 or 3 / 9 may be assigned to pets with easily palpable and somewhat visible ribs, pelvis, or spine, and / or severe waistline and abdominal reduction. A score of 2.5 / 5 or 4 / 9 may be assigned to pets with easily palpable, but not as visible, ribs, pelvis, or spine, and / or evident waistline RQfrC iP / ZZΖ / E / YILI and abdominal reduction. A score of 3 / 5 or 5 / 9 may be assigned to pets with palpable ribs, pelvis, or spine covered by a thin layer of fat, evident but not severe waist and abdominal reduction with more gradual curves, and / or minimal abdominal fat pad in front of the hind legs. A score of 3.5 / 5 or 6 / 9 may be assigned to pets with a light layer of fat over the ribs, spine, or pelvis making them more difficult to palpate, abdominal reduction present but minimal, a waist that is visible but not prominent, and / or minimal fat pad. A score of 4 / 5 or 7 / 9 may be assigned to pets with ribs covered with a thick layer of fat that requires finger pressure to feel, difficulty feeling the spine or pelvis, no apparent waist, abdominal reduction that is slightly visible, and / or moderate abdominal fat pad.A score of 4.5 / 5 or 8 / 9 may be assigned if the ribs, pelvis, and spine are covered with a dense layer of fat and palpable with extreme pressure, and / or there is no reduction in the abdomen or waistline. A score of 5 / 5 or 9 / 9 may be assigned if the ribs and spine are not palpable under a dense layer of fat, abdominal distension projects downward (which may be the opposite of abdominal reduction), the waistline is protruding or protruding, and there are fat deposits on the legs, face, or over the tailbone covering the pelvis. RQfrC ίη / ZZΖΠZ / E / YΙΛΙ with an extensive abdominal fat pad and sagging bellies. In some non-limiting modalities, a BCS of 4.5 or 7 / 9 may correlate with 30% body fat, or any other percentage of body fat.

[76] The age range of a dog, also referred to as the dog's life stage, can be categorized as young (5 years or less), middle-aged (greater than 5 years and less than or equal to 9 years), adult (greater than 9 years and less than or equal to 13 years), and senior (greater than 13 years). In some other non-limiting modalities, the dog's life stages can be categorized as puppy (newborn to reproductive maturity), junior (reproductively mature, still growing), adult (growth completed, structurally and socially mature), mature (from the middle to approximately the last 25% of life expectancy, which may be a window of time around the middle of the life expectancy for the breed), senior (from maturity to approximately the last 25% of life expectancy), and / or geriatric (at life expectancy and beyond).The cat's age range, also referred to as its life stage, can be kitten (from birth to 6 months), junior (7 months to 2 years), prime (3 to 6 years), mature (7 to 10 years), adult (11 to 14 years), and geriatric (over 15 years). The time period given for the number of visits can be 3 months, 6 months, 12 months, 24 months, etc. RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ months or any other period of time ranging from 0 to 144 months. In certain non-limiting modalities, kittens can be from 0 to 6 months, junior can be from 6 months to 2 years, full-grown can be from 3 to 6 years, mature can be from 7 to 10 years, and / or adult can be 11 years or more.

[77] The pet ownership dataset may include any information related to the pet owner, such as the number of cats or dogs owned (i.e., the number of cats or dogs owned by a given pet owner), the owner's age, any other consumer-related data, or any other demographic information related to the pet owner. Consumer-related data may include, for example, household income, home ownership, spending habits, geolocation, lifestyle attributes, weather patterns, environmental data, and / or household purchasing habits. In some non-limiting modalities, the physical or behavioral traits of a dog or cat dataset may include, for example, activity data, daily scratching minutes, nutrition consumption, bionomic data, and / or genetic and location data.One or more databases may include some overlap, meaning that there may be some common data included in multiple databases. RQfrC ίΠ / ΖΖηΖ / Ε / ΥΙΛΙ

[78] C. Linking data and predicting audiences or products

[79] As described above, in certain non-limiting modalities, data linking can be performed using identity resolution. Identity may refer to an anonymous consumer profile collected through one or more user devices. For example, constructing an individual's identity may involve combining data collected from laptop browsers, cell phones, email subscriptions, and / or offline purchases in a group. The combined data may then be anonymized so that marketers, or any other entity or person with access to the identity, do not have access to the consumer's personally identifiable information. Personally identifiable information, for example, may be a name, email address, telephone number, or postal address.In some non-limiting modalities, other consumer-related information, such as gender, age, location and / or consumer purchasing or browsing activities, may be included as part of the identity, which may be reviewed by marketers or any other entity or person with access to the identity.

[80] The resolution of identity may be the compilation RQfrC iP / ZZΖ / E / YILI of all the information in a profile. For example, identity resolution might indicate that a consumer conducts business activities in one browser but personal activities, such as personal purchases, in a different browser. In yet another example, identity resolution might indicate that the consumer's mobile device contains more data or better reflects the consumer's off-hours habits, such as preferred television programs or personal items searched for on one or more internet browsers. Identity resolution provides marketers, or any other entity or person with access to identity resolution, with a holistic view of the consumer.Identity resolution, therefore, can connect different identifiers across multiple platforms and / or devices to enable people-based targeting, personalization, or measurement. In some, but not limited to, modalities, identity resolution can be built or aggregated in real time.

[81] In other, non-limiting modalities, instead of using identity resolution, data linking can be performed based on device ID linking. Device ID linking can match a list of cookie IDs, such as pet care IDs, to one or more mobile device IDs, such as RQfrC iP / ZZΖ / E / YILI the Identifier for Advertisers (IDEA) or the Google or Android Advertising ID (GAID). A probabilistic device graph can be used as part of the device ID linking. In some non-limiting modalities, a deterministic graph can be used to match IDs obtained from multiple providers and indexed to the same IDs. When both the probabilistic and deterministic graphs indicate a device link, the device ID linking can be said to be an exact match or a true positive.

[82] In some non-limiting modalities, the probabilistic graph may indicate a link without the deterministic graph indicating a similar device link. Such modalities can occur when the deterministic graph is unaware of one or more device IDs included in the link. Since the deterministic graph may be unaware of one or more device IDs, the veracity of the link may be considered unknown and may be excluded from the device ID link analysis. In other non-limiting modalities, the deterministic graph may be aware of one or more device IDs, but those device IDs may be linked to IDs other than those shown as linked in the probabilistic graph. RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ modalities are known as inexact matching or false positives. When a link is neither on the probabilistic graph nor the deterministic graph, the link may be referred to as a potential link. The potential link may be counted as a true negative or may not be counted as part of the process or analysis.

[83] In certain non-limiting modalities, the deterministic graph may show a link between two or more devices, while the probabilistic graph does not show a similar link between the two or more devices. When the probabilistic graph is unaware of at least one of the device IDs known in the deterministic graph, the link may be considered unknown. On the other hand, when the probabilistic graph is aware of one or more devices but simply fails to link the devices, the link may be considered a false negative.

[84] Once the above quantities (e.g., true positive, false negative, false positive, or unknown linkage) are identified, the accuracy and range of the device identification linkage can be computed. Accuracy, for example, can be the rate of true positives to all positives, which can be calculated using the following equation: true positive + false negative. Range, for example, can be the rate of true positives to all positives. RQfrC ίη / ZZΖΠZ / E / YΙΛΙ positive to all true links, which can be calculated using the following equation: true positive (true positive + false negative^T-Ί Ί,Ί..,r &. The calculated measurement of accuracy and / or range using the above equations may be the lower bound on the true statistic. For example, the calculated accuracy may be approximately 90% and the range may be approximately 75%. In another example, the accuracy may vary from approximately 65% ​​to approximately 97%, while the range may vary from 20% to 80%. Increasing the accuracy by approximately 2-3% may reduce the range by approximately 50%.

[85] In some non-limiting modalities, linking may include associating or aggregating the consumer's multi-touch points. For example, a single consumer may have two email addresses, such as a personal email address and a business email address. To link the two email addresses, information or data from consumer database companies may be used. The consumer's full name, physical address, and / or email addresses may then be linked to a profile or identity resolution. The profile or identity resolution may also include other consumer attributes, for example, postal address, telephone number, or other relevant information. RQfrC ίΠ / ΖΖηΖ / Ε / ΥΙΛΙ cell phone, pet attribute, gender, and / or any other attribute.

[86] Figure 4 illustrates a linking system according to certain non-limiting modalities. In particular, Figure 4 may include data lake 410, which may be similar to data lakes 130, 240, and 340, and data consent framework 420, which may be similar to data consent frameworks 140, 250, and 350. The data included in one or more datasets can be leveraged to create a pet health care profile. The profile can be created using any known attributes about the pet. For example, the profile may include the following information: Pseudonym ID Breed Spend on VCA (in dollars) Health Plan Prescribed Product Pet's Age Pet BCS Pet Monitor 123124 Boxer $200 No Royal Canin 4 Ideal Yes 1241245 Pug $5,000 Yes Hills 1 Ideal Yes 251524 Bulldog $412 No Royal Canin 6 Ideal Yes 53552 Westie $512 Yes None 2 Ideal Yes 34233 Mixed Breed $514 Yes None 5 Ideal Yes Table 1. Example of a pet profile.

[87] Any known CDP or data management platform (DMP) can be used to create the one or more datasets and / or the pet healthcare profile. Creating the one or more datasets and / or the pet healthcare profile may include linking aggregated data from one or more sources or datasets. The healthcare profile may include one or more of the pet's healthcare needs. For example, a particular pet's healthcare profile may indicate that the pet has dental problems or poor dental hygiene. The 430 prediction tool can be used to predict pet products to help meet the pet's healthcare needs.For example, the 430 predictive tool can determine that a particular pet chew can help improve, resolve, or address a pet's dental problem, thereby meeting the pet's healthcare needs. In another example, the 430 predictive tool can help determine services, products, or medications based on one or more attributes. In certain non-limiting modalities, the predictive tool may utilize artificial intelligence or a machine learning algorithm.

[88] In some non-limiting modalities, the prediction tool may use a method or process called gradient reinforcement for machine learning. The method or process may be initiated with data from a single sector, such as pet monitoring data. Decision tree learning, which includes one or more decision trees, may then be used to model multiple scenarios to compare one computational method to another in a stepwise manner. RQfrC ίη / ZZΖΠZ / E / YΙΛΙ product to offer to the customer. Then, testing or modeling can be run on another sample to determine if the performance can be replicated without time-series data. For example, for a single scenario, media exposure + product purchase may produce 90% confidence in a purchase prediction, media exposure + product purchase + advertising response may produce 93.2% confidence in a purchase prediction, and / or media exposure + product purchase + advertising response + product purchase may produce 93.7% confidence in a purchase prediction. However, for a plurality of scenarios with different data entry points, product purchase may produce 70% confidence in a purchase prediction, product purchase + advertising response may produce 93% confidence.1% in a purchase prediction, and / or product purchase + response to advertisement + product purchase can produce a 93.4% confidence in a purchase prediction.

[90] Product purchase 435 and advertisement response 432 may include prior pet owner behaviors. These one or more inputs may be processed by the prediction tool 430, along with pet profile data, to determine a pet product or service to help meet the pet's healthcare needs. In other words, the tool RQfrC ίΠ / ΖΖηΖ / Ε / ΥΙΛΙ Prediction 430 can not only predict pet products or services to help meet a pet's healthcare needs, but also the purchasing behavior of a pet owner for a given pet. In some non-limiting applications, prediction 430 can also be used to identify an audience for advertising or marketing the pet product.

[91] Figure 5 illustrates a system or method for aggregation and linking according to certain non-limiting modalities. In particular, in step 510, data can be collected from one or more sources and aggregated into a data lake. The PII contained within the aggregated data can then be cleaned, meaning that the PII can be removed or separated from the data. In some non-limiting modalities, the cleaned data can be categorized into one or more datasets based on attributes. A pet profile can then be created based on the categorized data in the one or more datasets. A CDP, such as Epsilon, can be used to clean the PII from the aggregated data, to create the pet profile, and / or to categorize the one or more datasets. As shown in step 520, the aggregated data can be cleaned and / or reformatted to a unified, threshold, or standard quality level.After the data is at least one of the aggregated, cleaned, classified, and / or the pet profile is created, the data can be linked or connected. RQfrC iP / ZZΖ / E / YILI as shown in stage 530. Data can be linked or connected to an analytical model to perform relevant or pertinent tasks. For example, the marketing engine might use breed information, but the pet's pharmaceutical attributes are not relevant. In another example, supporting science might consider genetics and laboratory diagnostics instead of nutritional data. In stage 540, the linked data can be analyzed to predict the pet product based on aggregated or collected data to help meet the pet's healthcare needs. In certain non-limiting modalities, the data can be analyzed in a DMP. In stages 550 and 560, an advertisement for the pet product can be displayed on a screen or graphical user interface targeting the pet owner or audience member.In some non-limiting scenarios, stage 550 or 560 may include the use of a digital-side platform (DSP), including real-time bidding. Stage 550 may include the delivery of an analytics use case, while stage 560 may include the delivery of the advertisement to the pet owner or any other target audience. For example, stage 550 may be relevant because the output can be in summary form, without being aggregated. Stage 560 can be broken down to a pseudonym ID level to match other pseudonym IDs. RQfrC ίΠ / ΖΖηΖ / Ε / ΥΙΛΙ

[92] FIGURE 6A illustrates a system or method for aggregation and linking according to certain non-limiting modalities. In particular, FIGURE 6A shows a targeting process based on first-party data. First-party data, for example, can be collected from the health or location monitoring device 110, DNA or genetic tests 210, the POS device in a veterinary hospital 310, and / or any other first-party source. In certain non-limiting modalities, second-party or third-party data can also be used to accompany the first-party data. As shown in step 610, the first-party data can be extracted, aggregated, and / or mapped. The data, for example, can be extracted on a discrete basis, such as hourly, daily, weekly, and / or monthly. In step 611, the data can be aggregated, classified into datasets, and / or pet or pet owner profiles can be created.In some non-limiting modalities, step 611 can be performed by a CDP. For example, as shown in step 611, datasets can be matched and / or pet or pet owner profiles can be created by matching encrypted emails, device identification, and / or any other method. A pet or pet owner profile, in certain modalities, may be referred to as a pet care identifier.

[93] In stage 612, the identification of care of RQfrC ίΠ / ZZΖ / E / YΙΛΙ pets can be matched with a cookie or device ID and / or a DMP. In other words, step 612 can be used to update the DMP with data. In step 613, the DMP can allow a data audience to be created. Advertisements for pet products can then be delivered or streamed to the data audience, as shown in step 614. A target audience can be the audience selected to be best suited to see, read, or view the advertisement. The audience can include the pet owner or any other member for whom the advertised pet products might be relevant. In certain non-limiting modalities, step 615 can illustrate matching by having encrypted emails uploaded to a social networking site and / or a search engine.Then, advertisements can be delivered through a social networking site and / or a search engine.

[94] FIGURE 6B illustrates a system or method for aggregation and linking according to certain non-limiting modalities. In particular, FIGURE 6B illustrates a system or method for measuring the effectiveness of the system or method shown in FIGURES 5 and 6A. In step 620, a search engine and / or a social network may collect measurements or metrics on the advertising of the pet product, referred to as exposures, in FIGURES 5 and 6A. The measurements and / or metrics RQfrC ίη / ZZΖΠZ / E / YΙΛΙ can be aggregated in stage 621 and forwarded to data lake 622. Data lake 622 can be similar to data lakes 130, 240, 340, and 410. In certain non-limiting modes, monthly uploads of exposure files can be transmitted, as shown in stage 630. An exposure file can be non-aggregated, event-level data from an advertising campaign, typically reported by the publisher or media entity.The data may be raw data that can be collected from the advertising server logs, which may include, for example, the Internet protocol (IP) address, which represents the address of the user making the request, the user agent, which is a text string sent to the server by the browser that provides certain identifying information about the browser, the exposure time, the exposure date, the content of the advertisement, the publisher, the advertising unit, which may identify the location of the advertisement to be placed on the publisher's site, the line item, which may identify the respective client or creative campaign, the identification ranking, and / or other information related to that event.Other measures may include a pop-up blocker, a browser without Flash, a browser with JavaScript disabled or incapable, ad-blocking software, disabling image rendering, and / or automatic updates. RQfrC ίΠ / ΖΖηΖ / Ε / ΥΙΛΙ

[95] In stage 631, exposure files can be paired to a cookie or device ID by means of a CDP. From the CDP, the paired or updated data can be transmitted to data lake 632, which may be similar to data lake 622.

[96] FIGURE 7 illustrates a system or method for aggregation and linking according to certain non-limiting modalities. FIGURE 7 outlines a system for data ingestion, classification, processing for data, and then deploying those created audiences. In particular, FIGURE 7 illustrates the use of known 710 data and unknown 720 data. Known 710 data may include sales, services, marketing, commerce, and / or any other external data. Unknown 720 data may include web, mobile application, Internet of Things (IoT), media information, and / or any second-party or third-party information. Known 710 data and unknown 720 data may be collected and transmitted to the 730 pet activation platform. The 730 pet activation platform may be similar to the 140, 250, 350, and 420 consent frameworks shown in FIGURES 1 through 4.In some, non-limiting, modes, the 730 Pet Activation Platform may be cloud-based. The 730 Pet Activation Platform may use information that already complies with data privacy and data protection regulations, regardless of region. For example. Alternatively, or additionally, the 730 Pet Activation Platform can help target audiences for known and / or unknown customers through digital marketing. The 730 Pet Activation Platform can be used for digital experiences that can be personalized for one or more customers. In other, non-limiting ways, the 730 Pet Activation Platform can be used to connect with third-party advertising platforms.

[97] In certain non-limiting modalities, a DMP, also referred to as a unified data management platform (UDMP), can be a centralized system for collecting and / or analyzing large datasets originating from disparate sources. The 730 pet activation platform, for example, can be a DMP. A DMP can create a combined development and delivery environment that provides one or more users with consistent, accurate, and / or timely data. As described above, in its simplest form, a DMP could be a NoSQL database management system that can import data from many systems and / or allows marketers and publishers to view data in a consistent manner. Other DMPs, for example, can combine data management technologies and data analysis tools into a single software package, which may include an intuitive or easy-to-navigate executive dashboard.

[98] The DMP can be used to collect data RQfrC iP / ZZΖ / E / YILI structured and / or unstructured data from a variety of internal and external sources, and / or to integrate and store data. The DMP can also be used to analyze or organize data to provide data controller information for marketing and / or advertising campaigns. For example, data incorporated into a data management platform can be first-party data, second-party data, and / or third-party data. Third-party data can be used to fill gaps in a company's own data and partner data. In certain non-limiting modalities, any DMP known in the art can be used.

[99] In stage 740, the data can be personalized, meaning that pet profiles can be created, audiences can be determined, and / or pet healthcare needs can be predicted. Based in part on pet healthcare needs, advertisements can be automatically sent to known email addresses, mobile devices, search engines, social media, and / or via SMS to pet owners, as shown in stage 750. This may refer to a known trigger. Alternatively, or additionally, the advertisements can target an unknown trigger. An unknown trigger, for example, could be search ads, site ads, or other advertisements. RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ external advertising, social media advertising, and / or a designated pet care app or website. In stage 760, advertising-related measurements and / or metrics may be collected and / or transmitted.

[100] In certain non-limiting forms, the advertisement, for example, may comply with the Interactive Advertising Bureau (IAB) standards. The advertisement, for example, may be a fixed-size advertisement, a flexible-size advertisement, a native advertisement, and / or a lightweight advertisement. Advertisements may have various sizes, which may be based on the number of pixels on a screen that an advertisement can occupy. The pixel size, for example, may be the size of a billboard, the size of a smartphone ad, the size of a leaderboard, the size of a portrait, the size of a skyscraper, or a small, medium, or large feature phone ad.

[101] D. Results of the method and system described herein

[102] Figure 8 illustrates a diagram according to certain non-limiting modalities. In particular, Figure 8 can illustrate daily impressions and / or the performance of certain modalities described above. As shown in Figure 8, a daily impressions chart includes information RQfrC ίη / ZZΖΠZ / E / YΙΛΙ related to XYZ 810, pet care 820, and predictive pet care 830. XYZ 810, for example, can be a targeted audience based on purchases, built from an aggregation of the audience that bought a specific pet-related product. The pet-related product might have been purchased, for example, at a grocery store, pharmacy, wholesaler, e-commerce store, and / or specialty pet store. In certain, non-limiting ways, XYZ 810 can indicate a third-party audience partner and / or the results targeting the audience each day, showing the total number of viewable impressions for an advertisement. Pet care 820 can indicate the collective audiences used to achieve audience targeting and / or the total number of daily viewable impressions for the advertisement.Predictive pet care 830 can indicate collective audiences to achieve expanded audience targeting to include audiences similar to a base audience. The line can represent the total number of daily viewable impressions for advertisements.

[103] In some non-limiting modalities, predictive pet care 830 can illustrate daily impressions based on the system or method used in FIGURES 1 to 7. As shown in FIGURE 8, predictive pet care 830 RQfrC iP / ZZΖ / E / YILI may receive more daily impressions than Pet Care 820. Furthermore, FIGURE 8 also displays a bar chart showing the year-to-date performance of XYZ 810, Pet Care 820, and Predictive Pet Care 830. The visibility of NCS 810, Pet Care 820, and Predictive Pet Care 830 may be 78.70%, 80.30%, and 77.96%, respectively. The total conversion / matched impressions (VCR or CVR) of XYZ 810, Pet Care 820, and Predictive Pet Care 830, on the other hand, may be 71.6%, 75.8%, and 75.7%, respectively. This illustrates that the system or method shown in FIGURES 1 to 7 provides a significant technological improvement with enormous benefits over previously used systems.

[104] E. Flowchart of the method described herein

[105] FIGURE 9 illustrates a flowchart according to certain non-limiting modalities. In particular, FIGURE 9 illustrates a method 9000 implemented by a system shown in FIGURES 5 to 7 to meet a pet healthcare need. In step 910, the system may collect data from one or more disparate data sources. The one or more disparate data sources may include a health and location monitoring device, a DNA or genetic testing service, or a POS device in a veterinary hospital. The data may comprise at least one RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ of first data, second data, or third data. The collected data may include information about the pet's health. In step 920, the method may include aggregating the collected data from one or more disparate sources into a data lake. The method may also include removing personally identifiable information from the aggregated data in the data lake, as shown in step 930. In certain non-limiting modalities, the method may include classifying the removed data into one or more datasets based on one or more attributes of the pet or a pet owner. The one or more datasets may include at least one of the pet's health, pet ownership, or physical or behavioral traits.

[106] In some non-limiting modalities, the method may include reformatting data from one or more disparate data sources to a uniform level of quality. In stage 940, the method may include classifying the deleted data into one or more datasets based on one or more attributes of the pet or pet owner. The method in stage 950 may include creating a profile of a pet based on the data included in the one or more datasets. In stage 960, a pet product may be determined to help meet the pet's healthcare needs. Aggregation, deletion, classification, creation, and determination may be performed using a CDP or a DMP. In other non-limiting modalities, the method may include predicting a RQfrC iΠ / ZZΖ / E / YΙΛΙ Target audience for the pet product advertisement. The target audience may include the pet owner and one or more other pet owners. The determination or prediction may use a machine learning algorithm to predict the target audience or determine the pet product to help meet the pet's health care need.

[107] In certain non-limiting embodiments, the method may include addressing the pet owner with an advertisement for the pet product, as shown in step 970. The method may also include displaying the advertisement or an offer for the pet product to the pet owner through a terminal device graphical user interface, as shown in step 980.

[108] F. Diagrams of the system or apparatus described herein

[109] FIGURE 10 illustrates a diagram of a 10000 system according to certain non-limiting modalities. In particular, certain non-limiting modalities may include distributed resources that may include one or more hardware or cloud servers configured to perform any of the collection, processing, transmission, or storage described above in FIGURES 1 through 7. For example, in one modality of the disclosed subject matter, a 1000 system is provided. The 1000 system may comprise one or more RQfrC iP / ZZΖ / E / YILI components, such as one or more servers, which can collect data from one or more sources, data lakes, a data consent framework, and / or a prediction tool. With reference to the modality in FIGURE 10, for illustrative and non-limiting purposes, the system may include a central server 1100, which collects data from one or more individual servers 1200, 1300, 1400. In certain modalities, the central server 1100 may include a server computer, a desktop computer, a laptop computer, a cloud-based computing device, or other available computing devices. In some modalities, the central server 1100 may comprise multiple computers.

[110] Additionally, in certain configurations, the one or more individual servers 1200, 1300, 1400 may include a server computer, a desktop computer, a laptop computer, a cloud-based computing device, or other available computing devices. In this non-limiting configuration, the individual servers 1200, 1300, 1400 may collect data from the one or more distributed resources 1210-1230, 1310-1330, and 1410-1430, respectively, used to process data from a tested pet product. The distributed resources 1210-1230, 1310-1330, and / or 1410-1430 may be any imaging or non-imaging detector. In addition to static data, such as, but not limited to, the RQfrC ίη / ZZΖΠZ / E / YΙΛΙ Resource identification information including host or resource name, processor architecture (i.e., number of cores), and / or resource location; data collected from resources may include, but is not limited to, measurable information such as memory usage and / or availability (GB), CPU speed (MHz), and start and end times, such as the time when the resource is available or running.

[111] In the modality of FIGURE 10, for illustrative and non-limiting purposes, the collected data may be processed by at least one processing component, such as the 1600 processing server of the 1000 system, configured with logic to collect, analyze, process, and / or store data received from one or more resources. In certain modalities, the 1600 processing server may include a server computer, a desktop computer, a laptop computer, a cloud-based computing device, or other available computing devices. The processing server may comprise one or more processors contained within one or more separate servers, or alternatively, and as shown in dashed lines in the modality of FIGURE 10, it may be a standalone component configured, for example, to receive and transmit information to and from another server, such as a central 1100 server. RQfrC ίΠ / ΖΖηΖ / Ε / ΥΙΛΙ

[112] Based on the collected data, system 1000 can target an audience and / or a pet owner using a pet product advertisement. The advertisements can be viewed by the pet owner or an audience member on device 1500, the terminal. Device 1500, the terminal, can include a PC, a workstation, a user computer, and / or a mobile device. Consequently, in some modalities, the advertisement can be displayed in a graphical user interface.

[113] By way of example and not limitation, in one mode, data relating to one or more teeth of a pet may be collected from resources 1210-1230, 1310-1330, 1410-1430 distributed by one or more individual servers 1200, 1300, 1400, and stored in files on the individual servers 1200, 1300, 1400. The files may be any of a plurality of file types such as flat files, database files, markup language files, or the like. The central server 1100 of the mode described herein may receive two or more batch file transfers of the data collected from the individual servers 1200, 1300, 1400, respectively. Consequently, individual 1200, 1300, and 1400 servers can be specifically configured with memory and technology components of RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ file transfer to manage and transmit data to the central server over a network (not shown).

[114] FIGURE 11 is a diagram of an apparatus or system 1110 according to certain non-limiting embodiments. In particular embodiments, one or more computing systems 1110 may perform one or more stages of the method described or illustrated herein, such as those stages shown in FIGURES 1 through 9. In some embodiments, one or more computing systems 1110 provide the functionality of the server and / or computing device described or illustrated herein. In certain embodiments, software running on one or more computing systems 1110 performs one or more stages of one or more methods described or illustrated herein or provides the functionality described or illustrated herein. Particular embodiments include one or more portions of one or more computing systems 1110. Herein, reference to a computing system may encompass a computing device, and vice versa, where appropriate.In addition, the reference to a computing system may encompass one or more computing systems, where appropriate. The 1100 computing system may include one or more monitors, displays, and / or graphical user interfaces that allow pet owners and / or audience members to view advertisements for pet products.

[115] This disclosure contemplates any suitable number RQfrC iP / ZZΖ / E / YILI of 1110 computing systems. This disclosure considers the 1110 computing system to take any suitable physical form. By way of example, and not by way of limitation, the 1110 computing system may be an embedded computing system, a system-on-a-chip (SoC), a single-board computing system (SBC) (such as, for example, a computer-on-module (COM) or system-on-module (SOM)), a desktop computer system, a laptop computer system, an interactive kiosk, a mainframe computer, a network of computing systems, a mobile phone, a personal digital assistant (PDA), a server, a tablet computer system, an augmented / virtual reality device, or a combination of two or more of these.Where appropriate, the 1110 computing system may include one or more 1110 computing systems; it may be unitary or distributed; span multiple locations; span multiple machines; span multiple data centers; or reside in a cloud, which may include one or more cloud components in one or more networks. Where appropriate, one or more 1110 computing systems may perform, without substantial spatial or temporal limitation, one or more stages of one or more methods described or illustrated herein. By way of example, and not by way of limitation, one or more 1110 computing systems may perform in real time or in batch mode one or more stages of one or more methods described or illustrated herein. One or more 1110 computing systems may perform, at different times or in different locations, RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ one or more steps of one or more methods described or illustrated herein, where appropriate.

[116] In particular embodiments, the 1110 computing system includes a processor 1112, memory 1113, storage 1114, an input / output (I / O) interface 1115, a communication interface 1116, and a lili bus. Although this disclosure describes and illustrates a particular computing system having a particular number of particular components in a particular arrangement, this disclosure contemplates any suitable computing system having any suitable number of any suitable components in any suitable arrangement.

[117] In particular embodiments, the 1112 processor includes hardware for executing instructions, such as those that comprise a computer program. By way of example, and not as a limitation, to execute instructions, the 1112 processor may fetch instructions from an internal register, an internal cache, memory 1113, or storage 1114; decode and execute them; and then write one or more results to an internal register, an internal cache, memory, or storage 1114. In particular embodiments, the 1112 processor may include one or more internal caches for data, instructions, or addresses. This disclosure covers the 1112 processor that includes any suitable number of any suitable internal caches, where RQfrC iP / ZZΖ / E / YILI is appropriate. As an example, and not as a limitation, the 1112 processor may include one or more instruction caches, one or more data caches, and one or more translation-forward buffers (TLBs). Instructions in the instruction caches may be copies of instructions in memory or storage 1113, and the instruction caches may speed up the retrieval of those instructions by the 1112 processor. Data in the data caches may be copies of data in memory or storage 1113 for instructions executing the 1112 processor to operate on; the results of previous instructions executed by the 1112 processor to access, or write to, memory or storage 1113 by subsequent instructions executing the 1112 processor; or other suitable data. Data caches can speed up read or write operations by the 1112 processor.TLBs can accelerate virtual address translation for the 1112 processor. In certain configurations, the 1112 processor may include one or more internal registers for data, instructions, or addresses. This disclosure covers the 8020 processor, which includes any suitable number of any suitable internal registers, where appropriate. Where appropriate, the 1112 processor may include one or more arithmetic logic units (ALUs); be a multi-core processor; or include one or more 1112 processors. Although. RQfrC ίη / ZZΖΠZ / E / YΙΛΙ This disclosure describes and illustrates a particular processor; this disclosure covers any suitable processor.

[118] The 1112 processor can be implemented by any computing or data processing device, such as a central processing unit (CPU), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), digitally enhanced circuits, or a comparator device or a combination thereof. The processors can also be implemented as a single controller, or a plurality of controllers or processors.

[119] In certain configurations, memory 1113 includes main memory for storing instructions for processor 1112 to execute or data for processor 1112 to operate on. As an example, and not as a limitation, the 1110 computing system can load instructions from storage or another source (such as, for example, another 8000 computing system) into memory 1113. The 8020 processor can then load the instructions from memory 1113 into an internal register or cache. To execute the instructions, the 1112 processor can retrieve the instructions from the internal register or cache and decode them. During or after instruction execution, the 1112 processor can RQfrC iP / ZZΖ / E / YILI write one or more results (which may be intermediate or final results) to the internal register or internal cache. The 8020 processor may then write one or more of those results to memory 1113. In particular modes, the 1112 processor executes instructions only in one or more internal registers or internal caches or in memory 1113 and operates only on data in one or more internal registers or internal caches or in memory 1113 (as opposed to storage or elsewhere). One or more memory buses (which may each include an address bus and a data bus) may couple the 1112 processor to memory 1113. The lili bus may include one or more memory buses, as described below. In certain configurations, one or more memory management units (MMUs) reside between processor 1112 and memory 1113 and facilitate accesses to memory 1113 requested by processor 1112.In certain configurations, 1113 memory includes random access memory (RAM). This RAM may be volatile memory, where appropriate. Where appropriate, this RAM may be dynamic RAM (DRAM) or static RAM (SRAM). In addition, where appropriate, this RAM may be single-port or multi-port RAM. This disclosure covers any suitable RAM. 1113 memory may include one or more 1113 memories, where appropriate. Although this disclosure describes and illustrates particular memory, this disclosure covers any suitable memory.

[120] One or more memories 1113 may be shared, distributed, and / or hybrid shared-distributed. Shared memory architectures may be based on the global memory space, which may allow all nodes to share memory. In distributed memory architectures, on the other hand, processors have their own memory, and / or communication protocol and network to connect each compute node. Hybrid shared-distributed memory may include both shared and distributed memory architectures.

[121] In particular configurations, the 1110 computing system may also include storage. 1114 storage may include mass storage for data or instructions. By way of example, and not as a limitation, 1114 storage may include a hard disk drive (HDD), a floppy disk drive, flash memory, an optical disk, a magneto-optical disk, magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these. 1114 storage may include removable or non-removable (or fixed) media, where appropriate. 1114 storage may be internal or external to the 1110 computing system, where appropriate. In particular configurations, the storage may be non-volatile solid-state memory. In particular configurations, the storage may include read-only memory (ROM). Where appropriate, this ROM may be mask-programmed ROM, programmable ROM (PROM), RQfrC ίη / ΖΖΠΖ / Ε / ΥΙΛΙ Erasable PROM (EPROM), electrically erasable PROM (EEPROM), electrically alterable ROM (EAROM), or flash memory, or a combination of two or more of these. This disclosure covers mass storage in any suitable physical form. Storage 1114 may include one or more storage control units that facilitate communication between the processor 1112 and storage 1114, where appropriate. Where appropriate, storage may include one or more storage units. Although this disclosure describes and illustrates particular storage, this disclosure covers any suitable storage.

[122] In particular modalities, the 1115 I / O interface includes hardware, software, or both, providing one or more interfaces for communication between the 1110 computing system and one or more I / O devices. The 1110 computing system may include one or more of these I / O devices, where appropriate. One or more of these I / O devices may enable communication between a person and the 1110 computing system. By way of example, and not by way of limitation, an I / O device may include a keyboard, numeric keypad, microphone, monitor, mouse, printer, scanner, speaker, fixed camera, light pen, tablet, touchscreen, scroll wheel, video camera, other suitable I / O device, or a combination of two or more of these. The RQfrC iP / ZZΖ / E / YILI I / O device may include one or more detectors. This disclosure covers any suitable I / O devices and any suitable 1115 I / O interfaces for them. Where appropriate, the 1115 I / O interface may include one or more software drivers or devices that enable the 1112 processor to control one or more of these I / O devices. The 1115 I / O interface may include one or more 1115 I / O interfaces, where appropriate. Although this disclosure describes and illustrates a particular I / O interface, this disclosure covers any suitable I / O interface.

[123] In particular modalities, the 1116 communication interface includes hardware, software, or both that provide one or more interfaces for communication (such as, for example, packet-based communication) between the 1110 computing system and one or more other 1110 computing systems or one or more networks. By way of example, and not by way of limitation, the 1116 communication interface may include a network interface controller (NIC) or network adapter for communicating with an Ethernet or other wired network, or a wireless NIC (WNIC) or wireless adapter for communicating with a wireless network, such as a Wi-Fi network. This disclosure covers any suitable network and any 1116 communication interface suitable for it. By way of example, and not by way of limitation, the 1110 computing system may communicate with an ad hoc network, a personal area network (PAN), a local area network (LAN), a network of RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ wide area network (WAN), a metropolitan area network (MAN), or one or more portions of the Internet or a combination of two or more of these.

[124] One or more portions of one or more of these networks may be wired or wireless. As an example, the 1110 computing system may communicate with a wireless PAN (WPAN) (such as, for example, a Bluetooth WPAN), a Wi-Fi network, a WiMAX network, a cellular network (such as, for example, a Global System for Mobile Communications (GSM) network), or another suitable wireless network, or a combination of two or more of these. The 1110 computing system may include any 1116 communication interface suitable for any of these networks, where appropriate. The 1116 communication interface may include one or more 1116 communication interfaces, where appropriate. Although this disclosure describes and illustrates a particular communication interface, this disclosure contemplates any suitable communication interface. The communication interface may be used to receive and / or transmit information related to one or more of a pet's teeth.

[125] In particular configurations, the lili bus includes hardware, software, or both that couple components of the 1110 computing system to one another. By way of example, and not as a limitation, the lili bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, or a front-side bus. RQfrC iP / ZZΖ / E / YILI (FSB), a HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infiniband interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI Express (PCIe) bus, a Serial Advanced Technology Attached Bus (SATA), a Video Electronics Standards Association (VLB) local bus, or another suitable bus or a combination of two or more of these. The lili bus may include one or more 8120 buses, where appropriate. Although this disclosure describes and illustrates a particular bus, this disclosure contemplates any suitable bus or interconnect.

[126] Herein, a non-transient, computer-readable storage medium or media may include one or more semiconductor-based or other integrated circuits (ICs) (such as, for example, field-programmable gate arrays (FPGAs) or application-specific ICs (ASICs)), hard disk drives (HDDs), hybrid hard disk drives (HHDs), optical discs, optical disc drives (ODDs), magneto-optical discs, magneto-optical drives, floppy disks, floppy disk drives (FDDs), magnetic tapes, solid-state drives (SSDs), RAM drives, cards or drives, any other suitable non-transient, computer-readable storage medium, or any suitable combination of two or more of these, where appropriate. A non-transient storage medium, RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ computer readable, may be volatile, non-volatile, or a combination of volatile and non-volatile, where appropriate.

[127] The above methods provide significant advantages and technical improvements. In particular, the features described above provide data streamlining in a classified structure that allows multiple applications to run concurrently, rather than running a single application separately for each database. For example, running multiple applications concurrently may also be known as parallel processing of data-intensive or compute-intensive applications. Parallel processing can be classified as compute-intensive and / or data-intensive. Data-intensive applications face two main challenges, including processing exponentially growing volumes of data and significantly reducing data analysis cycles to enable timely decision-making.On the other hand, computationally intensive applications can be used to describe application programs that are heavily reliant on computing. Such computationally intensive applications may dedicate most of their execution time to computational requirements as opposed to input / output, and as such, they use relatively small volumes of data. In certain non-limiting cases, an application can be both data-intensive and computationally intensive. RQfrC ίη / ΖΖΠΖ / Ε / ΥΙΛΙ

[128] The classification between data-intensive and compute-intensive parallel processing can also be related to grid computing. A computing grid can be heterogeneous in nature, meaning that the grid can utilize different computing nodes and / or multiple geographically distributed nodes using wide area network communications. Grids can be used to solve complex computational problems that are compute-intensive. On the other hand, data-intensive computing systems can be homogeneous clusters, meaning that the nodes in the computing cluster can be identical and use local area communications between nodes. Grid computing can virtualize resources and provide the basis for cloud computing services. Consequently, cloud computing services can be used for data-intensive parallel computing.In other words, certain non-limiting modalities described above may utilize one or more cloud computing services. These cloud computing services may include a large pool of configurable virtual resources that can scale to accommodate varying parallel processing loads. Variable processing loads may include big data.

[129] The scope of this disclosure covers all changes, substitutions, variations, alterations RQfrC iP / ZZΖ / E / YILI modifications to the example modalities described or illustrated herein that would be understood by a person of ordinary skill in the art. The scope of this disclosure is not limited to the example modalities described or illustrated herein. Furthermore, although this disclosure describes and illustrates the respective modalities herein as including particular components, elements, features, functions, operations, or steps, any of these modalities may include any combination or permutation of any of the components, elements, features, functions, operations, or steps described or illustrated anywhere herein that would be understood by a person of ordinary skill in the art.Additionally, any reference in the appended claims to an apparatus or system or a component of an apparatus or system that is adapted, arranged, capable of, configured, enabled, operable, or operative for performing a particular function encompasses that apparatus, system, or component, whether or not it or that particular function is activated, turned on, or unlocked, provided that such apparatus, system, or component is so adapted, arranged, capable, configured, enabled, operable, or operative. Furthermore, although this disclosure describes or illustrates particular embodiments as providing particular advantages, the particular embodiments may provide none, some, or all of these advantages.

Claims

CLAIMS 1. A method for satisfying a pet health care need comprising: collecting data from one or more disparate data sources; aggregating the collected data from the one or more disparate data sources into a data lake; deleting personally identifying information from the aggregated data in the data lake; classifying the deleted data into one or more datasets based on one or more attributes of the pet or a pet owner; creating a profile of a pet based on the data included in the one or more datasets; determining a pet product to help satisfy the pet health care need; and targeting the pet owner with an advertisement for the pet product.

2. The method according to claim 1, further comprising: displaying the pet product advertisement to the pet owner through a graphical user interface of a terminal device.

3. The method according to claim 1 or 2, wherein the data comprise at least one of first data, second data, or third data.

4. The method according to any of claims 1 to 3, wherein the one or more disparate data sources comprise a health and location monitoring device, a DNA or genetic testing service, or a point-of-sale device in a veterinary hospital.

5. The method according to any of claims 1 to 4, wherein the collected data comprises information on the pet's health.

6. The method according to any of claims 1 to 5, wherein the aggregation, deletion, classification, creation, and determination are performed by means of a consumer data platform or a data management platform.

7. The method according to any of claims 1 to 6, further comprising: reformatting data from one or more disparate data sources to a uniform level of quality.

8. The method according to any of claims 1 to 7, further comprising: predicting a target audience for the pet product advertisement, wherein the target audience may include the pet owner and one or more other pet owners. RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ 9. The method according to any of claims 1 to 8, wherein the determination or prediction uses a machine learning algorithm for predicting the target audience or determining the pet product to help meet the pet's health care need.

10. The method according to any one of claims 1 to 9, wherein the one or more data sets comprise at least one of the pet's health, pet's ownership, or the pet's physical or behavioral traits.

11. The method according to any one of claims 1 to 10, wherein the one or more disparate data sources are first-party data collected from a health and location monitoring device.

12. The method according to any of claims 1 to 10, wherein the one or more disparate data sources are one or more genetic testing services.

13. The method according to claim 12, wherein the data lake includes a data consent framework.

14. The method according to any of claims 1 to 13, wherein the data consent framework comprises a transparency and consent chain.

15. The method according to any of claims 8 to 14, wherein the prediction further comprises an analysis of covariance, comprising the steps of: assessing whether a plurality of means of a dependent variable are equal across a plurality of levels of a categorical independent variable; controlling for a plurality of effects of a plurality of other variables; wherein the analysis of covariance is used in either a pre-test or post-test analysis.

16. A system for satisfying a pet health care need comprising: at least one processor; at least one memory comprising computer program code; and wherein the computer program code is configured, when executed by the at least one processor, to cause the system to: collect data from one or more disparate data sources; aggregate the data collected from the one or more disparate sources into a data lake; erase personally identifiable information from the data aggregated into the data lake; classify the erased data into one or more datasets based on one or more attributes of the pet or a pet owner; create a profile of a pet based on the data included in the one or more datasets; and determine a pet product to help satisfy the pet health care need.and target the pet owner with an advertisement for the pet product.

17. The system according to claim 16, wherein the computer program code is configured, when executed by at least one processor, to cause the system to: display the pet product advertisement to the pet owner through a graphical user interface of a terminal device.

18. The system according to claim 16 or 17, wherein the data comprise at least one of first data, second data, or third data.

19. The system according to any of claims 16 to 18, wherein the one or more disparate data sources comprise a health and location monitoring device, a DNA or genetic testing service, or a point-of-sale device in a veterinary hospital.

20. The system according to any of claims 16 to 19, wherein the collected data comprises information on the pet's health. RQfrC ίΠ / ZZΖηZ / E / YΙΛΙ 21. The system according to any of claims 16 to 20, wherein the aggregation, deletion, classification, creation, and determination are performed by means of a consumer data platform or a data management platform.

22. The system according to any of claims 16 to 21, wherein the computer program code is configured, when executed by at least one processor, to cause the system to: reformat data from one or more disparate data sources to a uniform level of quality.

23. The system according to any of claims 16 to 22, wherein the computer program code is configured, when executed by at least one processor, to cause the system to: predict a target audience for the pet product advertisement, wherein the target audience may include the pet owner and one or more other pet owners.

24. The system according to any of claims 16 to 23, wherein the determination or prediction uses a machine learning algorithm for predicting the target audience or determining the pet product to help meet the pet's healthcare needs. RQfrC iP / ZZΖ / E / YILI 25. The system according to any of claims 16 to 24, wherein the one or more data sets comprise at least one of the pet's health, pet's ownership, or the pet's physical or behavioral traits.

26. The system according to any of claims 16 to 25, wherein the one or more disparate data sources are first data collected from a health and location monitoring device.

27. Systems according to any of claims 16 to 25, wherein the one or more disparate data sources are one or more genetic testing services.

28. The system according to any of claims 16 to 26, wherein the data lake includes a data consent framework.

29. The system according to claim 28, wherein the data consent framework comprises a transparency and consent chain.

30. The system according to any of claims 23 to 27, wherein the computer program code is configured, when executed by at least one processor, to cause the system to perform an analysis of covariance configured to: evaluate whether a plurality of means of a dependent variable are equal across a plurality of levels of a categorical independent variable; control for a plurality of effects of a plurality of other variables; wherein the analysis of covariance is used in either a pre-test or post-test analysis.

Images