WO2022103307A1

WO2022103307A1 - Universal identifier for mobile advertising

Info

Publication number: WO2022103307A1
Application number: PCT/RU2021/050371
Authority: WO
Inventors: Мария Давидовна ГОРЬКОВА
Original assignee: Общество С Ограниченной Ответственностью "Квает Медиа"
Priority date: 2020-11-12
Filing date: 2021-11-08
Publication date: 2022-05-19
Also published as: RU2755812C2; RU2020137257A3; RU2020137257A

Abstract

Proposed is a universal identifier for mobile advertising, which constitutes a method for preventing the storage and circulation of the consumer profile of a mobile cellular subscriber by third-party domains in that the mobile cellular operator creates a plurality of data structures. Each data structure consists of a plurality of pairs of values of a subscriber identifier and a hash value, and each data structure is mapped to a network address of a web server, wherein the subscriber identifier is any standardized subscriber identifier (MSISDN, IMSI, IMEI), and the hash valve HASH is the value of the subscriber identifier converted by a hash function. The claimed method includes integrating QIT Script program instructions into a program logic for processing HTTP queries from partner web servers. An operator responds to a network HTTP query from a partner web server, initiated by QIT Script program instructions, by transmitting a HASH value that corresponds to a network address of a web server and to the subscriber's MSISDN.

Description

Mobile advertising universal identifier.

The present invention relates to the field of telecommunications and communication, preferably to mobile marketing systems, any types of Internet advertising and audience data processing systems, and can be used in cellular networks of GSM, UMTS, IMS standards.

It is known that the effectiveness of mobile advertising increases significantly if sales of web inventory (places for advertising placement) by web resources (web sites) are accompanied by an accurate marketing portrait of the subscriber who visited the web-pecypc. The subscriber's marketing portrait includes data on the thematic interest of the subscriber, his age, gender, and other data that allow him to display potentially interesting advertising content. The collection and processing of this data is provided by DMP (Data Management Platform) systems based on the subscriber's behavior in the network - gender, age, browsing history, search, geo data, etc. A large amount of such characteristics provides web resources with the ability to more accurately recognize the marketing portrait of a subscriber who has visited the site. This means increasing the cost of advertising space in technological platforms for the sale of web inventory SSP (Supply / Sell Side Platform) and its price for an advertiser participating in an OpenRTB advertising auction or direct sales (Direct Binding) on the side of technological platforms for the purchase of web inventory DSP (Demand Side platform). The uncontrolled circulation of this, although anonymous data, nevertheless potentially influencing the behavior of the subscriber in the network, has led to a global trend to limit their circulation.

Well-known anonymous identifiers used in the Internet advertising industry to analyze audience data in order to form personalized advertising to the subscriber - cookies, in particular the so-called. third-party cookies containing anonymous information about the subscriber, mobile device identifiers GAID (Google Advertising ID) and IDFA (Identifier for Advertisers) of Android and IOs operating systems. The data associated with these identifiers allows Internet marketers to understand the thematic interest of an anonymous subscriber, measure the effectiveness of advertising campaigns, calculate the ratio (conversion) of ad impressions to clicks, app installs, etc.

The trend of strengthening the protection of personal data on the Internet directly affects the technology of online advertising. As of 2022, Google will stop supporting third party cookies, Apple has already stopped supporting cookies from third party domains. Mobile device developers are tightening organizational and technical measures, counteracting uncontrolled tracking of data on mobile devices of subscribers. So GAID and IDFA have ceased to be stable since 2018 and can be changed by the subscriber. As of 2020, IDFA is disabled by default, and applications are required to request subscriber consent to activate it.77

Major players in the Internet industry offer their solutions aimed at a compromise between the confidentiality of subscriber data and the need to track his preferences - Google Privacy Sandbox, Apple SKAdNetwork. Such initiatives can lead to undesirable dominance of large companies and monopolization of the online advertising market, which is economically harmful to the industry as a whole.

Under the current conditions, Internet marketers use indirect user identification techniques (the so-called browser fingerprinting) based on the settings and features of its equipment - screen resolution, fonts, MIME types, encryption modes, etc. Identifiers are used that are not directly intended for mobile marketing systems - IP and MAC addresses, mobile device address Serial Number. Some of them are known for instability and / or the need to obtain permission from the subscriber to use. All these methods are united by one drawback - the impossibility of providing protection against uncontrolled circulation of subscriber data in the network. So it is known that many sites and mobile applications generate a significant part of their income by selling audience data to large aggregators (third-party domains).

The ubiquitous availability of mobile devices leads to an increase in mobile traffic. According to various estimates, its growth from 2017 to 2020 was about 30%. The cellular network operator NP (Network Provider) assigns unique standardized identifiers MSISDN, IMSI, IMEI to each subscriber, and the services provided by the operator must be associated with them (clause 1.5 of the ETSI TS 100 616 specification). It is known that, starting from the UMTS architecture, cellular operators everywhere use software and hardware systems for deep traffic analysis DPI (Deep Packet Inspection) to prioritize GPRS traffic at the SGSN/GGSN node. The analysis is necessary to manage the limited bandwidth of the communication channel and charge the subscriber for the consumption of services. To use GPRS packet data services in cellular networks, the PDP Context procedure is performed, which includes assigning a static or dynamic IP address to a mobile device of an MSISDN subscriber. AAA (Authentication, Authorization, Accounting) protocols are used to describe access to services and control their consumption. Taking into account these conditions of the current state of the art, the authors propose a universal method for identifying a subscriber, excluding unwanted circulation of his marketing portrait and guaranteeing a stable identifier for the mobile advertising industry. The universal identifier proposed by the authors can be used to evaluate advertising campaigns and process audience data instead of identifiers offered by major participants in the online advertising industry. A significant economic effect of the method proposed by the authors lies in the impossibility of major players in the Internet industry to dominate the Internet advertising industry.

It is known that mobile operators, by definition, process the personal data of their home subscribers. Therefore, the processing of anonymous data that forms their marketing portraits associated with hashed identifiers, which, in essence, the proposed method is impossible to identify by third-party domains, ensures the impossibility of their circulation. On the other hand, a sufficient number of cellular operators, including virtual ones, implies a sufficient level of competition within the proposed identification method and a corresponding economic effect.

The closest technical solution adopted for the prototype is the application US2020126106 dated 12/17/2019. describing a way to anonymously identify a user in two different user spaces. The essence of the method (US2020126106 fig.1) is to create hash tables (8A, 8B) with subscriber data in each user space, and create in one of these spaces a module for mapping the hash value of one space into the hash value of another (mapping module 12) . The disadvantage of this solution is that the user hash value (HASH) is identical for different user spaces. Thus, the proposed method performs anonymous user authentication in different user spaces, since the identifiers (UID) are different for different spaces, and the display module is included in only one user space. However, the method does not prevent user identification between spaces, and hence the transmission of a hash value (HASH), behind which the user is hidden, to a third-party web resource for accumulating data about his behavior and subsequent circulation of this data. As the person skilled in the art will understand, the method extends to an unlimited number of such spaces with the same drawback. It is noted that the method of American authors does not aim to counteract the exchange of the hash value of the user identifier between spaces, but rather supports the exchange of user information. The stated goal and technical result of the invention proposed by the authors is to prevent the accumulation and circulation of data of the marketing portrait of a cellular network subscriber by third-party domains and guaranteed provision of relevant advertising content to the subscriber based on his marketing portrait. Such a goal, according to the authors, is achievable provided that on a network node (QIT) independent of user spaces (web resources) a universal identifier UID (Universal Identifier) is created, which is a hash value of the hash function to the subscriber identifier used in cellular networks (MSISDN, IMSI, IMEI) for connection switching and service billing. Moreover, the hash function for each web resource is different in terms of ensuring the impossibility of displaying this hash value between any two different web resources.

The technical result is achieved due to the universal mobile advertising identifier, which includes the creation by the mobile operator of a plurality of data structures, each data structure of which consists of a plurality of pairs of subscriber identifier and hash values, the establishment of each data structure corresponding to the network address of the web server, and upon receipt by the network node SGSN / GGSN or any other network node of the cellular operator of a network request from the network address of the web server, the cellular operator defines a hash value in a set of data structures that corresponds to the network address of the web server and any standardized MSISDN, IMSI, IMEI subscriber identifier used in networks cellular communication for connection switching and billing of services, and transmits this hash value in the network response to the network address of the web server. In this case, the data structure is any associative array that stores and processes a set of pairs of subscriber identifier values and a hash value obtained by converting this subscriber identifier with a hash function. The network address of the web server is the URL address of any server on the Internet, and the network request from the network address of the web server is an HTTP request to the web server initiated by the execution of program instructions of the scripting programming language pre-installed on the web server by the mobile operator.

In FIG. 1 The general scheme of interaction of network nodes within the framework of the proposed method of universal identification of a subscriber UID is presented.

In FIG. 2 A generalized exchange diagram of network nodes is presented when executing the UID universal subscriber identification method. The figures do not limit the network architecture, the sequence of calls or the interaction scheme of entities within the framework of the proposed invention, but illustrate one of the preferred options that provides the claimed technical result.

In the general case, the following definitions are accepted in the context of the application: MSISDN - the subscriber's mobile device identifier and the subscriber's mobile device MS (Mobile Station) are interchangeable. Third-party web resource or third domain - a web site or domain on the Internet with which the subscriber's mobile device does not establish a web session via the HTTP protocol in a certain, limited period of time. Marketing portrait of a subscriber - any anonymized data of a subscriber that provides recognition of his marketing characteristics and thematic interest associated with any digital identifier. Turnover of the subscriber's marketing portrait - accumulation, sale, transfer of subscriber's anonymous data to third-party web resources or third domains.

The proposed universal mobile advertising identifier UID can be implemented on any GSM, UMTS, IMS cellular network architecture. The UID management software logic can be executed on any nodes that are part of the packet cera PS (Packet Switched Domain clause 3.3.2 GPP TS 23.002) of the cellular operator NP. Including on the main node PS - SGSN / GGSN which performs the functions of managing packet communication sessions. In the preferred embodiment of FIG. 1.2 The software logic of the UID is executed in the memory of the QIT server, which receives all GPRS traffic from the SGSN/GGSN node, for example through a software bypass.

The QIT server hardware is implemented on X86 architecture servers running a Unix-like operating system. The QIT software architecture is the know-how of the applicant. The QIT programming algorithm involves creating a set M of URL m.. (URI m) data structures, each of which implements a key-value associative array interface. In the general case, this is any data structure that performs programmatic operations of writing, searching, reading, deleting a key by value. In a particular implementation of Figure 1, the URL data structure m.. is a standard hash table. Hash table keys - a set of N unique identifiers of MSISDN N subscribers that are used to process connections on standard network nodes (SSP, SCP, SGSN, GGSN, HLR, etc.) of the CN switching system (Core Network clause 3.2.3 GPP TS 23.002 standard ) operator NP cellular. In general, the keys may be MSISDN, IMSI, IMEI, or any other stable identifier, preferably standardized by ITU-T and 3GPP and associated with connection switching. and service billing (i.1.5 ETSI TS 100 616 specification). Hash table values are hash values obtained by applying a hash function to associated keys. The software algorithm uses standard hash functions known from the prior art, including the use of hash function modifiers, etc. cryptographic salt.

Determining the MSISDN of a subscriber by his web session IP address can be performed using the PDP Context data, which is a standard procedure performed in cellular networks for servicing a subscriber in GPRS packet mode. The PDP Context procedure includes allocating an MSISDN n subscriber's mobile device an IP address and activating an APN (Access Point Name). Based on the IP address of the web session and the time of assignment of this MSISDN IP address to the mobile device, the QIT software algorithm can determine the MSISDN n which, unlike the IP address, is a stable identifier due to the provisions of the telecommunications standards ITU-T, 3GGP and ETSI. Correspondence to the current IP of the subscriber and MSISDN can be established by the QIT program logic using AAA (Authentication, Authorization, Accounting) protocol messages. AAA protocols (RADIUS, DIAMETR) contain data on the correspondence between the MSISDN of the subscriber and the IP assigned to him for processing packet network data. Access to PDP Context data and AAA protocol interfaces is provided by software interaction between QIT and SGSN/GGSN.

Set of N pairs of MSISDN values n, m; HASH n, m of each data structure URL m is a subset of all subscribers of the operator (figure 1). Economically, the set N in each URL data structure m may be equal to the set of all subscribers of the operator, for example, if the operator's advertising policy NP provides for the application of the same advertising rules for all subscribers. Or not equal, for example, if the operator NP implements the management of advertising web content by subscription. Without limiting the proposed UID method, it is noted that in private implementations, the subset N in different URL data structures m may not be equal, based on the marketing preferences of the NP operator.

For each data structure URL m, the QIT software algorithm maps the network address of the web server URL m. A new data structure URL m is created upon detection by the QIT program logic of a web session with a network address of the URL of the web server that is not included in the set M. In this case, the QIT program algorithm assigns the new network address of the URL m of the web server to the new data structure and writes to it the MSISDN value associated with the discovered web session. At the same time, the QIT software algorithm does not create a hash value HASH n for the newly recorded MSISDN n the newly created URL m data structure, since the web server does not consider affiliate URL. The newly created URL m data structure is updated with new MSISDN n as web sessions MSISDN n are found in the URL address (FIG. 1 URL data structure M-1). The QIT programming logic can implement different rules for creating a new URL m data structure, depending on the marketing preferences of the operator.

Technically, a web server at the network address URL m is any standard network node that provides access to a set of web pages via the HTTP protocol, for example, a news resource (Fig.1 URL ml, URL m2, URL M-1). Economically, the web server at the network address URL m may be a partner of the NP operator (FIG. 1 URL ml, URL m2), or not be a partner of the NP operator (FIG. 1 URL M-1). In the case of a partnership agreement with a web resource at the network address URL m, QIT integrates into the software architecture of the web server URL m a set of program instructions executed in a scripting programming language known from the prior art, for example, JavaScript ( 1 QIT Script). This set of instructions initiates an HTTP call from the web server URL w to the QIT network address to obtain the HASH of the subscriber of interest, and may include additional logic, such as requesting the subscriber's marketing profile IDs (in terms of Internet advertising segment IDs). QIT Script may include any other logic provided by the partnership agreement. At the same time, the QIT software algorithm defines the web server URL m as a partner and applies a unique hash function to all MSISDN keys n,m stored in the data structure of the new partner URL m to generate hash values HASH n,m.

If an HTTP POST request initiated by the QIT Script partner web server URL ml is detected, the software algorithm responds with an HTTP GET with the HASH ml parameter and segment identifiers if the QIT Script program instructions provide for a request for subscriber segment identifiers (marketing portrait). The affiliate web server uses the subscriber's marketing portrait in the RTB (Real-Time Binding) advertising auction or direct sales to its DB (Direct Binding) partners, stimulating the increase in the price of its web inventory. A network request from a non-partner web server will fail because the QIT programming logic will not determine the data structure associated with the network address of the web server URL w and HASH with the value of the corresponding MSISDN n. In case the third party web server copies the QIT Script from the web server partner URL ml or m2 then triggered by the copied QIT Script a network request to the QIT server from this third party web server will also fail. Since the QIT program logic will not determine the data structure of the URL and HASH of the subscriber corresponding to the address of the third-party web server and the MSISDN of the subscriber. Such QIT software logic reliably prevents the accumulation and circulation of subscriber marketing portrait data by third-party web resources.

A generalized network exchange of network nodes within the framework of the method proposed by the authors is shown in Fig.2. To simplify figure 2, a generalized network dialogue is shown on the example of the interaction of QIT with a partner web server at the network address URL ml. Without loss of generality of reasoning, interaction with other partner web servers (m2, etc.) is performed similarly. In figure 2, it is assumed that the software architecture of the web server URL ml includes QIT Script, at the appointed time time N-1 PDP Context MSISDN N-1 is activated, the subscriber's mobile device MSISDN N-1 is assigned an IP address and the mobile device operates in receive mode GPRS packet data.

The web browser of the mobile device of the MSISDN N-1 subscriber establishes a web session with the web server URL ml and requests the web page index.html of the message 1-3 URL ml GET index.html. The URL ml web server performs standard request processing 3, a QIT Script is triggered that initiates a 4,5 POST (IP) request to the QIT address, in the parameters of which the web server transmits the current IP of the subscriber of the established web session. QIT processes the request 5, determines the subscriber's MSISDN (determ MSISDN N-1) by the IP of the subscriber and the time of its destination (detect IP, time N-1), determines the HASH N-1, ml (determ HASH N- 1, ml) and responds with messages 6,7 GET7HASH N-1, ml. Web server URL ml accepts HASH N-1, ml and generates a request to the SSP platform to execute the RTB auction algorithm for the sale of its web inventory 8 API (inventory, HASH N-1, ml.). The SSP platform generates a request to QIT to obtain segment identifiers for the HASH N-1, ml value of the message 9,10 REQ API (HASH N-1, ml). The QIT processes the request and sends the subscriber's marketing portrait (segment IDs) to the SSP message 11,12 RESP API (segment ID). After that, the SSP launches the standard RTB auction logic RTB start up with an accurate marketing portrait of the MSISDN N-1 subscriber (segment ID). At the same time, the web server sends to the MSISDN N-1 web browser the index.html web page of message 13-15 2000k index.html, which the web browser renders to the index.html downloading subscriber. When processing the index.html web page, the web browser generates a request to the web server URL ml to download the advertising content of the message 16-18 js-code cellbanners REQ. After polling advertisers (DSP platforms in figure 2 is not shown), the SSP determines the advertiser, which sends its advertising material to the web server for downloading 19 APIs (cellbanners) and completes the RTB complete advertising auction. On request 18, the web server sends the received advertising material for downloading (cellbanners downloading) to the user's web browser MSISDN N-1 messages 20-22 js-code cellbanners RESP.

As a specialist understands, instead of holding an RTB auction, the URL ml web server can perform direct sales of DB (Direct Binding) web inventory to partners. In this case, the advertiser's DSP of the UML ml web server partner will also form a network request not to a third-party DMP, but to the QIT address using HASH N-1, ml. It is also obvious that the method does not lose generality in the case of a web server contacting an advertising network to sell web inventory. The method allows sending the subscriber's marketing profile (segment ID) to the partner's web server URL ml in messages 6,7 instead of HASH N-1, ml. Then the URL ml web server sends to the SSP platform not the HASH N-1, ml value, but the segment IDs. In this case, the SSP has a complete marketing portrait of the subscriber for holding an RTB auction, and the network exchange 9-12 SSP-QIT for his request becomes redundant.

In any case, as disclosed above, the method proposed by the authors unconditionally ensures the secrecy of the marketing portrait of the subscriber due to the circulation of the HASH N-1 value, unique for the UML ml web server, ml, which is impossible for accumulation and circulation by third-party web resources. A unique HASH value excludes:

- accumulation and circulation of the subscriber's marketing portrait by third-party DMPs;

- Appeal of the SSP of the platform to third-party DMPs to obtain a marketing portrait of the subscriber;

- obtaining a marketing portrait of a subscriber by third-party web resources by means of non-standard copying of QIT Script from the web server of the operator's partner NP.

At the same time, the method provides for linking the marketing profile to a stable MSISDN identifier, the stability of which is determined by the technical specification. For additional secrecy of the marketing portrait of a subscriber based on HASH and to prevent the accumulation of his statistics by third-party web resources, the QIT algorithm can provide for a periodic change in the hash function for any subset of the sets M and N.

Claims

The formula The universal identifier of mobile advertising, which includes the creation by the mobile operator of a set of data structures, each data structure of which consists of a set of pairs of subscriber identifier and hash values, the establishment of each data structure corresponding to the network address of the web server, characterized in that when received by the network node of the operator cellular network request from the network address of the web server, the cellular operator determines in a set of data structures a hash value that corresponds to the network address of the web server and the subscriber identifier used in cellular networks for switching connections and billing services, and passes this hash value to network response to the network address of the web server. The universal mobile advertising identifier according to claim 1, characterized in that the data structure is any associative array that stores and processes a set of pairs of subscriber identifier values and a hash value obtained by converting this subscriber ID hash function. A universal mobile advertising identifier according to claim 1, characterized in that the identifier used in cellular networks for switching connections and billing services is any standardized identifier MSISDN, IMSI, IMEI of the subscriber. The universal mobile advertising identifier according to i.1, characterized in that the cellular operator's network node is the SGSN/GGSN node or any other network node that performs the functions of network packet switching and traffic analysis. The universal mobile advertising identifier according to i.1, characterized in that the network address of the web server is the URL address of any server on the Internet, and the network request from the network address of the web server is an HTTP request of the web server initiated by the execution of program instructions of the scripting programming language in advance installed on the web server by the mobile operator.