Advanced data and analysis management platform

Abstract

The present disclosure relates to an advanced data management platform, method, and system that allow multiple users to remotely access a centralized database and analyze and extract predicted information, trends, etc. from the dataset using analytical tools, models, and algorithms. Further, the data management platform, method, and system include options that allow data analysis and model training with anonymized data. This in turn allows the trained model to be applied to a data set outside the data platform and / or system, thereby deriving predictive information about the data set.

Description

Technical Field

[0001] The present invention relates to an advanced data and analysis management platform and also to a system and method for implementing the advanced data and analysis management platform. Background Art

[0002] In many traditional organizations, standard practice is to purge data after a period of time, or at best retain the data in a traditional data warehouse for long-term storage and basic reporting.

[0003] In recent years, some organizations have begun to approach data in a more rigorous and advanced way, influenced at least in part by the latest buzzwords, success stories, and hot trends in publications and conferences. In the process, people have invested heavily in various "analytics" solutions or tools, but in many cases, they have made arbitrary decisions.

[0004] During this transition phase, from both an organizational and technological perspective, different teams and departments within an organization or group are likely to join the above “movement” - some more conservative, some quite enthusiastic, and many of them will invest in data and analytics tools through their own vendors (traditional and emerging).

[0005] As a result, fragmented data and analytics resources may be created, rather than a holistic, enterprise-wide, end-to-end technology approach (especially in large organizations). Some of the drawbacks of this fragmented system include siloed data storage, suboptimal basic analytical resources, and uncoordinated and incompatible actions, which lead to poor results, waste or duplication of organizational budget and resources, and missed opportunities. In addition, these haphazard and unconsulted accumulation of different approaches and countless assets continue to create new organizational silos and / or exacerbate existing ones. There is a lack of a single "source of truth" and a single "data point", which also applies to analytical repositories and reference materials. For example, different teams within the same organization may define and implement a single key performance indicator (KPI), process or use case differently. This in turn leads to, for example, different ways of reporting the same KPI and different values ​​(numbers) communicated throughout the organization and externally (for example, to investors and regulators). This may have an adverse impact on the performance, image, governance and / or benchmarking of the organization.

[0006] These distributed, unrelated, and uncoordinated analytical methods are "unrelated to each other" and are committed to achieving common effects and goals within the organization, but they cannot work together. Due to the cannibalization effect, push-pull effect, etc., the actual results are unsatisfactory and even counterproductive.

[0007] Another key point in this regard is the lack of the right kind and amount of data for data science activities. Analytics tend to be most effective when large amounts of data of different types are combined with rigorous analytical techniques. In large organizations, individual departments or functions may rely on other departments or functions to perform the services that the entire organization provides to customers. Similarly, for data and analytics, the fusion of different data and external data points can help realize the true analytical potential of the organization.

[0008] Without the above, a siloed approach without a centralized, dedicated data and analytics platform can result in poor efficacy and predictive power for analytical assets, such as machine learning (ML) models. This results in low accuracy and impact in predictions and decisions. Without a single point of development for correct and complete data, nor an associated ecosystem and single point of production, many analytical assets may remain on personal machines (such as local laptops) and fail to deliver benefits to the organization in production (either in the business or on IT systems). This naturally leads to a lot of duplication and waste, as well as the other issues mentioned above.

[0009] It is found that a common challenge in leveraging an organization’s big data assets to drive business value is often not simply deriving good actionable insights from machine learning models, but how to apply them. Building machine learning models into production business systems is often where data science teams fail (rather than developing the models themselves), and most resources may be focused on this.

[0010] In view of the above, there is an urgent need for an end-to-end overall technical solution that can at least partially solve the above technical problems. The inventor tried to find a solution in the public domain based on his experience in multiple fields and industries, but did not find such a "ready-made" tool or solution. This is not surprising, because creating such an end-to-end solution requires a variety of requirements, components, processes and services.

[0011] Although research has been conducted in this area and some entities have developed products / services that address certain aspects of the problem, such as Google's products "Colossus" and "Bigtable", the existing functionality is limited and cannot provide the required comprehensive technical solution. In particular, the applicant has not been able to find an end-to-end holistic analysis cycle solution that allows countless components, services, processes and skills to interact harmoniously and is vendor-independent. In order to meet the above needs, the inventor has developed the technical solution of the present invention. Summary of the invention

[0012] The present invention generally provides an advanced data and analysis management platform. Aspects of the present invention may include a platform operation method and a system that forms a part of or is coupled to the platform (see the following content and / or the accompanying drawings).

[0013] For ease of reference, the acronym "ADAM (Advanced Data and Analytics Management)" is used below and in some of the accompanying drawings to refer to the advanced data and analytics management platform.

[0014] The ADAM platform is configured to provide a single centralized source of development, management, production, repository and monitoring for all analytical assets (particularly advanced analytical tools) for an entire organization (e.g., an organization comprising one or more groups of companies). As shown in the figure, in the following example, the organization is MTN, including its operating company (OpCo). OpCos may be located in multiple countries, and in some countries, there may be only one OpCo for each country of business operation. It should be understood that the reference to MTN is for illustration only, and the ADAM platform can be applied to different industries and organizations. In the case of MTN, the local / field data system may be referred to as an "EVA" system, so reference may be made to the EVA system integrated with the ADAM platform.

[0015] The ADAM platform provides a data science workbench and asset management platform to make the organization's data development capabilities more mature. The ADAM platform can help popularize data science within the organization, promote local talent development, and enhance business flow empowerment.

[0016] Preferably, the platform should be centrally located, easily and securely accessible from anywhere, have the appropriate kind and amount of data (subject to legal review), have a complete ecosystem of analytical resources, be connected to every operating company in the group, and be able to create, reuse, industrialize (automate) data or analytical resources across the entire group, at operating company sites, in the cloud or at the edge (such as shopping malls or airports, etc.), or for business-to-business clients.

[0017] Essentially, the ADAM platform can be configured as an analytics factory and / or analytics-as-a-service platform, such as a one-stop collaboration shop, analytics code repository, model catalog, CICD (continuous integration and continuous development) pipeline for important organizational data sources (points), analytical assets, and supporting analytical algorithms, languages, etc.

[0018] The ADAM platform may include one or more of the following components / features:

[0019] · A single logical bit provides ready access to large amounts of carefully organized internal and external data;

[0020] · Support various analysis tools and languages;

[0021] · Access a library of modeling algorithms, review them, and download them in a secure, on-premises environment;

[0022] ·Ability to support various analytics (e.g. AI / ML / DL / heuristics / simulation, etc.);

[0023] · Connect different organizational entities, such as Group and OpCo data source systems;

[0024] · Ability to upload data, download model code, and upload model performance again for monitoring (e.g. via a CICD pipeline);

[0025] · Easily productize (industrialize) to automatically and regularly generate actionable insights, insight-based actions, and / or predictions.

[0026] The platform empowers users (from data scientists to basic analysts) to connect to and access parts of a broad enterprise-level infrastructure from anywhere, from which they can use their preferred technologies, programming languages, and analytical algorithm libraries. When using it, users such as data scientists only need to perform minimal settings to directly connect to data sources in the central data lake and improve business productivity.

[0027] Therefore, taking ML / AI models as an example, the platform provides a technical solution that can facilitate the “unleashing” of ML / AI value in large organizations.

[0028] Embodiments of the present invention can be extended to one or more computer program products for implementing the ADAM platform, which include at least one computer-readable storage medium, in which the implemented program instructions can be executed by at least one computer to enable at least one computer to implement basically the above-mentioned technology and realize basically the above-mentioned features.

[0029] The computer readable storage medium may be a non-transitory storage medium. The computer program product may be implemented in a variety of devices and locations, etc.

[0030] The ADAM platform can be any suitable computer or server, or include any suitable computer or server. The computer ADAM platform can be implemented in a distributed cloud computing environment, where remote processing devices linked by a communication network perform tasks. In a distributed cloud computer environment, the program modules executed by the ADAM platform can be located locally or remotely.

[0031] According to a first aspect of the present invention, there is provided a data management platform, comprising: a centralized database storing benchmark data, the centralized database being configured to receive one or more data sets from one or more data sources, the centralized database allowing one or more users to remotely access the centralized database; an analysis module being communicatively coupled to the centralized database, the analysis module comprising one or more analysis tools allowing data analysis of the one or more data sets; a model module being communicatively coupled to the centralized database, the model module comprising one or more algorithms for executing a set of instructions on the one or more data sets, the model module being capable of being trained by the one or more algorithms by comparing the one or more data sets with benchmark data, thereby deriving one or more output data sets having predictive information about the one or more data sets.

[0032] The one or more data sources may be one or more data warehouses that may be located in one or more countries. The one or more data sets may include customer data.

[0033] The data management platform can be configured to allow one or more users to remotely access a centralized database via a network connection to transfer one or more data sets from one or more data sources to the centralized database, wherein the network connection allows one or more users to analyze the one or more data sets using an analysis module including an analysis tool, and the network connection allows one or more users to train a model module including one or more algorithms by comparing the one or more data sets with benchmark data, thereby deriving an output data set having predictive information about the one or more data sets.

[0034] An analysis module including one or more analysis tools and a model module including one or more algorithms can be dynamically updated and evolved, wherein each instance of the one or more analysis tools analyzes one or more data sets, and the model module is trained by one or more algorithms by comparing the one or more data sets with benchmark data, thereby deriving one or more output data sets having predictive information about the one or more data sets.

[0035] Dynamic updating may refer to uploading data in real time.

[0036] Evolutions can be operational improvements to analysis modules that provide improved data analysis and model modules that are able to provide faster and more powerful predictive information when applied to one or more data sets.

[0037] The benchmark data may be dynamically updated by comparing each instance of the one or more data sets to the benchmark data.

[0038] Training a model module comprising one or more algorithms allows for faster export of each subsequent output dataset containing predictive information, including additional predictive insights about the one or more datasets.

[0039] The data management platform may further include a processing module for processing the data into one or more vectors.

[0040] One or more vectors can be broken down into one or more use cases.

[0041] Each use case can be configured for a specific purpose.

[0042] Each purpose can be to use a vector for a specific type of data.

[0043] The one or more vectors are stored as a repository on a data management platform, where one or more users can remotely access the repository to facilitate applying a model module including one or more algorithms to one or more data sets.

[0044] The availability of such one or more vectors stored as a repository on a data management platform would allow the application of a model module comprising one or more algorithms to one or more datasets that may be located in one or more countries.

[0045] The availability of the one or more vectors stored as a repository on the data management platform allows a model module including one or more algorithms to be applied to the one or more data sets with low latency compared to transmitting the one or more data sets to the data management platform.

[0046] The network connection may be a secure network connection, a private network connection, or a combination thereof.

[0047] The one or more data sources may be one or more data warehouses that may be located in one or more countries.

[0048] The centralized database may be configured to receive only one or more anonymized data sets, thereby deriving anonymized data.

[0049] The one or more analytical tools may be selected from the following group: Tableau, Oracle Business Intelligence, IBM Cognos Analytics, SAS, Microsoft Power BI, Amazon Redshift, Google BigQuery, Snowflake, Alteryx, Cloudera, Apache Hadoop, Google Vertex AI, Microsoft Azure Synapse, Microsoft Data Explorer, CosmosDB, Redis, Azure Cognitive Services, Azure Machine Learning, Spark, Databricks, Sqream, Confluent Kafka, Presto, Trino, Flare, HIDS, and combinations thereof.

[0050] The one or more algorithms may be selected from the group consisting of: machine learning algorithms, artificial intelligence algorithms, deep learning algorithms, heuristic algorithms, and combinations thereof.

[0051] One or more data sets may include customer usage information selected from the group consisting of customer voice call usage, customer screen time usage, customer data usage, customer messaging usage, customer device hardware specifications, customer transactions, customer interactions, customer behavior, customer revenue, network operations, network usage, network investment, internal operations, sales information, distribution information, agent operations, merchant operations, agent services, merchant operations, business-to-business products, business-to-business product services, digital products, OTT applications, customer value management, pricing, operations management, portfolio management, customer location, network location, network transmission, network configuration, network security, and combinations thereof.

[0052] The information contained in one or more data sets may be selected from the following group of customer information about customer usage: voice, internet, data, payment, digital services, enterprise services, network solutions, and combinations thereof

[0053] Predictive information may include information about customer behavior, thereby deriving user-specific product offerings.

[0054] A user-specific product offering may, for example, offer a customer one or more products associated with the model module and the predictive information derived for the customer by the one or more algorithms.

[0055] More specifically, the user-specific product offering may, for example, recommend to a customer one or more products including voice minutes, data, or a combination thereof.

[0056] The predictive information may include information indicating a preferred communication format to increase the likelihood that the customer will purchase the product offering.

[0057] The data management platform may include a data transmission module for transmitting the model module to a database located outside the centralized database, wherein the model module including one or more algorithms can be applied to one or more data sets available on the database to derive an output data set having predictive information about the one or more data sets.

[0058] The data transfer module allows the trained model module and one or more algorithms to be stored on a computer-readable storage medium.

[0059] The computer readable storage medium may be a non-transitory storage medium.

[0060] The computer-readable storage medium may also be a remote storage, which may include one or more of the following instances or units: cloud storage, a remote server, multiple processors communicatively coupled to each other, a computer, and combinations thereof.

[0061] The model module can be uploaded to any device or medium that includes one or more data sets, and then applied to the one or more data sets.

[0062] Uploading model modules to a location that can be outside of the data management platform would allow for reduced latency.

[0063] Uploading a model module to a location that can be located outside of a data management platform allows the model module to be applied to one or more datasets in a location from which the removal of one or more datasets would violate one or more data regulations.

[0064] Uploading the model module to a location that may be located external to the data management platform allows the model module trained on the anonymized data on the data management platform to export predictive information about one or more data sets at the location, which can be used to derive user-specific product offerings for one or more customers contained in the one or more data sets at the location.

[0065] The data management platform may also include an anonymous module that may be located outside the centralized database, and the anonymous module includes: a data module, used to receive one or more data sets from one or more data sources; an anonymization algorithm, used to anonymize the one or more data sets to derive anonymized data; a transmission module, used to transmit the anonymized data to the centralized database.

[0066] The anonymized data may be used to train a model module comprising one or more algorithms by comparing the anonymized data with benchmark data stored in a centralized database, thereby deriving an output data set having prediction information about the anonymized data.

[0067] A model module comprising one or more algorithms may be configured to be trained using anonymized data, wherein the model module is applied to one or more datasets located in a selected location to comply with country-specific data protection and privacy regulations.

[0068] The data management platform may be communicatively coupled to an external data source storing one or more data sets, which may be located external to the platform.

[0069] One or more data sets stored on the data source may be anonymized by the anonymization module to derive anonymized data. A model module including one or more algorithms may be trained by the anonymized data receivable by the centralized database, and the trained model module including one or more algorithms may be applied to one or more data sets that may be located at a data source (which may be located outside the centralized database).

[0070] The data management platform may also include an interface module that is communicatively coupled to the model module to receive prediction information about the one or more data sets, and the interface module is also communicatively coupled to one or more user devices to allow the user devices to access the prediction information about the one or more data sets.

[0071] According to a second aspect of the present invention, there is provided a method for training a model with anonymized data and applying the model to one or more data sets located at a selected location, the method comprising the following steps: providing a database storing benchmark data, the database being capable of receiving data from one or more data sources; providing an external data source communicatively coupled to the database, anonymizing one or more data sets stored in the external data source by executing a set of instructions on the one or more data sets stored in the external data source, thereby deriving anonymized data; using one or more analysis tools to allow data analysis to be performed on the data stored in the database; training a model comprising one or more algorithms by comparing the anonymized data with the benchmark data; and applying the trained model to one or more data sets located outside the database, thereby deriving predictive information about the one or more data sets.

[0072] The method may include the steps of storing the model on a storage medium, allowing the model to be applied to one or more data sets in any source communicatively coupled to the storage medium.

[0073] The method may include the step of using the forecast information to derive user-specific product offers.

[0074] According to a third aspect of the present invention, a digital management system is provided, comprising: a computing device including a processor, the processor being communicatively coupled to a memory, the memory being capable of storing one or more data sets obtained from one or more data sources, the memory storing benchmark data; an analysis module including an analysis tool, the analysis module being communicatively coupled to the memory, wherein the processor is capable of executing a set of instructions to enable the analysis tool to perform analysis operations on the one or more data sets; and a model module including one or more algorithms, the model module being communicatively coupled to the memory, wherein the model module is capable of being trained by the processor by applying one or more algorithms by comparing the one or more data sets with benchmark data, thereby deriving an output data set having predictive information about the one or more data sets.

[0075] The memory may be selected from the group consisting of: a non-transitory storage medium, a transient storage medium, and a combination thereof.

[0076] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, wherein the instructions stored in the medium, when executed by a computer, cause the computer to execute a method of using a computer system to train a model with anonymized data and apply the model to data at a selected location, the method comprising the following steps: providing a database storing benchmark data, the database being capable of receiving data from one or more data sources; providing an external data source communicatively coupled to the database, anonymizing one or more data sets stored in the external data source by executing a set of instructions on one or more data sets stored in the external data source, thereby deriving anonymized data; using one or more analysis tools to analyze the data stored in the database; training a model including one or more algorithms by comparing the anonymized data with the benchmark data; and applying the trained model to an external data set that may be located outside the database, thereby deriving predictive information about the external data set. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] The following further illustrates the specific implementation of the present invention with reference to the accompanying drawings.

[0078] Figure 1 The following is a schematic diagram showing how the ADAM platform can be integrated into an organization's enterprise data system.

[0079] Figure 2 A schematic diagram showing how the ADAM platform interacts with anonymized data (personally identifiable information or "PII" data) available through the EVA system.

[0080] Figure 3 A schematic diagram showing how the ADAM platform can be deployed as a “group platform”.

[0081] Figure 4 is an exemplary data system architecture including an example ADAM platform.

[0082] Figure 5 A diagram of the data pipeline between the ADAM platform and field systems is provided.

[0083] Figure 6 An overview of the operation of an exemplary ADAM platform is provided.

[0084] Figure 7 A table specifically lists an exemplary platform that complies with the first set of architectural principles.

[0085] Figure 8 A table specifically lists an exemplary platform that complies with the second set of architectural principles.

[0086] Fig. 9 A table specifically lists an exemplary platform that complies with the third set of architectural principles.

[0087] Fig.10 FIG. 4 is a schematic diagram of an exemplary logical architecture that can be used in an embodiment of the present invention.

[0088] Fig.11 A schematic diagram showing a first option for integrating a corporate instance of ADAM with operating companies within an organization.

[0089] Fig.12 A schematic diagram showing a second option for integrating a corporate instance of ADAM with operating companies within an organization.

[0090] Fig.13 A block diagram of an exemplary computer system capable of executing a computer program product to provide functions and / or actions according to aspects of the present invention. DETAILED DESCRIPTION

[0091] The following content is intended to explain the technical solution of the present invention, which is only to illustrate the relevant principles of the present invention, and is not intended to limit the scope of protection of the present invention. The embodiments in the text and the drawings can be changed, but the technical effects of the present invention can still be obtained and / or without departing from the scope of protection of the present invention. It should also be understood that some of the end effects or advantages of the present invention can be achieved by selecting certain features of the present invention without other features. Therefore, those skilled in the art will recognize that the present invention can be modified and adjusted (even more desirable in some cases), which can constitute an integral part of the present invention.

[0092] Embodiments of the present invention provide a "one-stop service", that is, a centralized platform for end-to-end analysis of asset lifecycle management. Such assets are, for example, ML models.

[0093] Figure 1A schematic diagram showing how the ADAM platform is integrated into an organization's enterprise data system is shown as an example. Data is received from various sources and stored in various formats such as text, video, images, etc. Once the data is stored, it can be utilized through a series of operations as shown in the "Organise&Consume" area. These operations then convert the data into a suitable format, so that the ADAM platform can interact with the data. This series of operations and conversions of the data are performed by each participant or contributor of the entire ADAM platform, which is called an EVA system as mentioned above. The key point is that each location or entire country will contribute one or more unique data sets as an EVA system.

[0094] Once the data is properly stored, the ADAM platform can perform further operations on it. These operations include but are not limited to analysis, visualization, modeling, training of data sets through ML, DL, AI, heuristic algorithms and other algorithms, application of ML, DL, AI, heuristic algorithms and other algorithms to one or more data sets for analysis and related predictions.

[0095] An important feature of the ADAM platform is that it is centralized, allowing the platform to be exposed to and trained on a variety of data sets received from a variety of sources and departments within one or more organizations. In this way, similar and unique data sets can be paired, integrated and / or "cross-fertilized", allowing the ADAM platform to make unique predictions that are not possible in platforms or models trained primarily on similar data sets.

[0096] Figure 2 A schematic diagram showing how the ADAM platform interacts with anonymized data (personally identifiable information or "PII" data) available through the EVA system. First, the EVA system obtains raw data from each location or country. Second, the EVA system anonymizes the raw data by performing a series of operations. Third, the anonymized data is moved or transmitted from the location or country to a cloud server that centrally hosts the ADAM platform. Fourth, the format of the anonymized data still allows the ADAM platform to use the anonymized data for query and training. Fifth, the ADAM platform and its algorithms and / or models are deployed to various locations or countries to analyze and predict real data, which is different from the conventional operation of sending or transmitting data to other media that can interact with it. Sixth, due to the availability of relevant data sets, the analysis and predictions for each location or country are unique, allowing each EVA system (location or country) to monetize these results.

[0097] This is an important feature of the present invention, as exporting data from each country is likely to violate or breach data regulations, such as GDPR or POPI. Therefore, using anonymized data to train the ADAM platform and related models, and then employing the ADAM platform and related models on real data in each location to obtain analysis and make predictions, allows the use of a large data lake without violating user privacy.

[0098] Figure 3 The following is a schematic diagram showing how to deploy the ADAM platform as a "group platform". Figure 2 As described above, the platform can be connected to the operational company data system via CICD. Such advantages include retaining intellectual property and ownership, asset reusability, and a "Build Once Deploy Many" strategy.

[0099] In the embodiment of the present invention, the ADAM platform is intended to prevent the above problems and shortcomings from occurring, and promotes the process of developing models, passing tests and putting into production. If it were not for the technical solution such as ADAM, organizational resources would have to build such systems, such as using open source solutions (based on individual or team notebook computers), which are easy to obtain but often have major challenges in terms of reliability and scalability. Such models also require a lot of work to maintain their operability.

[0100] The typical estimated time to productize an ML model (through such individual open source technologies and methods) may exceed 3 months, while in an embodiment of the ADAM platform, with the same resource pool, the estimated time may be reduced to, for example, 2 to 3 weeks. This becomes critical considering that today's enterprises need a large number of such models to increase revenue. Therefore, the modeling team within the organization can use the platform as shown in the figure to meet the needs of the organization.

[0101] Some key steps in the platform management implementation may include:

[0102] · Leverage governance committees

[0103] · Innovation Center, Collaboration

[0104] · Internal cloud resources

[0105] · End-to-end lifecycle of analytical assets (e.g. models) - DevOps / MLOpS, up to AiOp creation, comparison, agile delivery, productionization, measurement, retraining / rebuilding, etc.

[0106] Some of the platform’s features / modules may include:

[0107] · Analyze asset catalog

[0108] · Analyzing the codebase

[0109] · Self-healing model

[0110] · Management Model

[0111] · Central development model management and collaboration environment

[0112] · Decision Science Workbench - not only for ML, DL, but also for simulation, prediction, AI, etc.

[0113] · Decentralized development and deployment - Make Once, Use Many

[0114] · Compliant with organizational data warehouse (e.g. MTN EVA / DaaS)

[0115] · Batch or frequency models and streaming or real-time processing and modeling

[0116] · Access model algorithms, packages, and runtime libraries

[0117] · Model interoperability

[0118] · implement

[0119] · decision making

[0120] · rule

[0121] · suggestion

[0122] In the embodiment of the present invention, various users can interact with the ADAM platform:

[0123] · Data Scientist

[0124] · Coding tools - e.g. R, Python, SAS Miner, etc.

[0125] · Model / Release Manager

[0126] ·Consumers - such as insight analysts, Excel assistants, business analysts, enterprise and IT systems and processes, external B2B partners, etc.

[0127] Figure 4 An exemplary data system architecture including an example ADAM platform. The exemplary ADAM platform includes access to data belonging to the Operations Support System (OSS) and Business Support System (BSS) software systems, which are typically used by telecommunications and other service providers to manage operations and support business functions. In addition, the ADAM platform allows visualization and import of data from existing and external data sources. Multiple APIs are also integrated and available immediately. The ADAM model can then be trained and deployed on a variety of data sets, including large data lakes that continuously receive additional data sets, thereby more effectively training and deploying any subsequent data sets of the ADAM platform and model.

[0128] Figure 5 A data pipeline diagram between the ADAM platform and field systems is provided. As mentioned above, field data for each location or country is anonymized before being transferred or uploaded to the ADAM platform through continuous integration / continuous delivery (CI / CD). Once trained, the ADAM platform can be deployed to any location by simply downloading the model code for real data.

[0129] Figure 6 An operational overview of an exemplary ADAM platform is provided.The Business Semantic Layer (BSL) includes the curated and managed data locations where business users will consume generated metrics and KPIs.

[0130] Figures 7 to 9 Specifically lists examples of how the ADAM platform conforms to a set of architectural principles.

[0131] Fig.10 FIG. 4 is a schematic diagram of an exemplary logical architecture that can be used in an embodiment of the present invention.

[0132] Table 1 below summarizes the layers of an exemplary architecture blueprint, indicating the overall responsibilities that each component layer of the architecture may assume to implement the ADAM platform.

[0133]

[0134] Table 1

[0135] The ADAM platform can be divided into two logical parts: the control plane and the user plane. The control plane can be a centralized plane, such as at the group level in an organization, where most ADAM activities are run and available to all connected OpCos. In some cases, the control plane may only have virtual or anonymized data. Some OpCos may develop outside of ADAM or import models into ADAM. In the user plane, any models or other analytical assets developed on the ADAM platform are deployed to the OpCo (or other local / field locations) to run locally, and the control plane can be configured to monitor the performance and accuracy of the model. For example, for entities in embargoed countries, the developed models can be run locally without connecting to ADAM's control plane.

[0136] Fig.11 and Fig.12 Two example options for integrating the ADAM platform into the OpCo entity are shown, again using MTN as an example. Essentially, the first option uses a fiber network ring and network nodes as the network topology, and then uses the .NET software framework and Multiprotocol Label Switching (MPLS) as the routing technology. Fig.12 The second option in CI / CD uses CI / CD as a software development methodology that helps improve the speed and efficiency of the software development process by building, testing, and deploying software, building and testing code changes as they are committed to the code repository. Essentially, these options

[0137] Table 2 below summarizes some example functions / modules of the ADAM platform.

[0138]

[0139]

[0140] Table 2

[0141] The platform caters to collaboration, internal crowdsourcing and asset utilization and can be configured to allow the deployment and execution of analytical resources in the cloud, at the edge and on-site locally (e.g. at the group headquarters office, any operating company, affiliated units, customer locations and / or public locations around the world).

[0142] The ADAM platform can also:

[0143] · Integrate internal and external customers, partners such as operating companies, and group functional departments

[0144] (CVM, Fintech, Sales, Corporate, Network, Technology, Public (Government) and Private Partnerships).

[0145] · By enhancing or improving existing ways of working and use cases, new intelligent AI / ML analytical technology solutions (such as real-time insights, alert-based actions and data-driven decisions) are provided to enable and enrich all business lines within and outside the organization.

[0146] · Improve productivity and efficiency of data scientists, business analysts, non-technical staff, and senior management.

[0147] · Develop an organizational modernization strategy by changing your approach to data and analytics and creating an environment that is conducive to the spread of data and analytics benefits.

[0148] One or more of the above techniques may be implemented in or using one or more computer systems, such as Fig.13 The computer system 100 shown. The computer system 100 may be or include any suitable computer or server. The computer system 100 may be implemented in a distributed cloud computing environment, where remote processing devices linked via a communication network perform tasks. In a distributed cloud computing environment, program modules executed by the computer system 100 may be located locally or remotely.

[0149] Fig.13 In the example shown, computer system 100 has the characteristics of a general purpose computer. These components may include, but are not limited to, at least one processor 102, memory 104, and bus 106, which couples various components of system 100 (including memory 104) to processor 102. Bus 106 may have any suitable type of bus structure. Computer system 100 may include one or more different types of readable media, such as removable media and non-removable media and volatile and non-volatile media.

[0150] Thus, the memory 104 may include volatile memory 108 (e.g., random access memory (RAM) and / or cache memory), and may also include other storage media, such as a storage system 110 configured to read / write from / to a non-removable non-volatile medium (such as a hard drive). It should be understood that the computer system 100 may also include or be coupled to a magnetic disk drive and / or an optical disk drive (not shown) for reading or writing suitable non-volatile media. These media may be connected to the bus 106 via one or more data media interfaces.

[0151] The memory 104 may be configured to store program modules 112. The modules 112 may include, for example, an operating system, one or more application programs, other program modules, and program data, all of which may include implementing a networking environment. The components of the computer system 100 may be implemented as modules 112, and generally perform the functions and / or methods of the embodiments of the present invention as described herein. It should be understood that embodiments of the present invention may include or be implemented as multiple computer systems 100 that may be communicatively coupled to each other.

[0152] The computer system 100 may be operatively communicatively coupled to at least one external device 114. For example, the computer system 100 may communicate with external devices 114 in the form of a modem, keyboard, and display. These communications may be accomplished via appropriate input / output (I / O) interfaces 116.

[0153] The computer system 100 may also be configured to communicate with at least one network 120 (eg, the Internet or a local area network) via a network interface device 118 / network adapter. As described above, the network interface device 118 may communicate with other components of the computer system 110 via the bus 106.

[0154] Reference Fig.13 The components shown and described are examples only, and it should be understood that other components may be used in place of or in combination with those shown in the figures.

[0155] Working Example

[0156] Refer to the above Figures 1 to 13 Based on the teachings of the present invention, the present invention can be illustrated by examples of applying the present invention in various OpCos of MTN.

[0157] EVA systems in multiple countries generate and / or acquire large datasets related to customer behavior, consumption habits, etc. Each EVA system then anonymizes this data and uploads it to the ADAM platform. The ADAM model performs various models, algorithms, operations, and other functions on the anonymized datasets. The key point is that in each instance where the ADAM model is trained with an anonymized dataset, regardless of the similarity of that dataset to any other dataset, as the ADAM model is continuously trained, the ADAM model is able to analyze and make predictions on each anonymized dataset. This is one of the unique features of the centralized ADAM model; it is not only trained with homogeneous datasets, but also continuously improved with unique anonymized datasets provided by OpCo.

[0158] Once the ADAM model is trained, the model code can be downloaded and deployed on each EVA system for use on real data.

[0159] Specifically for MTN OpCo, the ADAM model is deployed as follows after training:

[0160] - Predict that customers are likely to be inactive within X days so that personalized marketing campaigns can be carried out for these customers;

[0161] -Divide customers into different market segments so as to carry out targeted marketing campaigns for these customers;

[0162] - Identify trends in customer data usage so that we can conduct customized marketing activities for customers in a timely manner;

[0163] -Identify whether customers are at home, in a business, hospital, or other organization and whether customers are members of the same organization in order to conduct targeted marketing activities for these customers.

[0164] The above are examples only, but it is important to note that datasets from different OpCos in different industries and / or with different functions allow for the identification of important trends that would otherwise not be identified if the model was simply using a homogenized dataset.

[0165] Other considerations

[0166] Aspects of the present invention may be implemented as systems, methods and / or computer program products. Therefore, aspects of the present invention may take the form of hardware, software and / or a combination of hardware and software, collectively referred to herein as "components", "units", "modules", "systems", "elements", etc.

[0167] It should be understood that the modules involved in the embodiments of the present invention are a set of codes or software programs that can perform one or more specific tasks. It should also be understood that each module includes multiple software programs or codes that can perform the same or similar functions.

[0168] Communicative coupling shall refer to, but not be limited to, the exchange of information, commands or data between devices or platforms.

[0169] In addition, aspects of the present invention may take the form of a computer program product embodied in one or more computer-readable storage media having computer-readable program code embodied thereon. The computer-readable storage medium may, for example, be an electronic, magnetic, optical, electromagnetic, infrared or semiconductor system, device or apparatus, or any suitable combination thereof. In the context of this specification, a computer-readable storage medium may be any suitable medium capable of storing a system, device or apparatus executing or associated programs. The program code / instructions may be executed by a single device, by multiple devices (e.g., local and remote devices), as a single program or as part of a larger system / software package.

[0170] The present invention may be implemented on any suitable form of computer system, including a stand-alone computer or a processor participating in a computer network. Therefore, a computer system programmed with instructions for implementing the disclosed method and / or system, a computer system programmed to perform various aspects of the present invention, and / or a medium storing computer-readable instructions (for converting a general-purpose computer into a system based on various aspects of the present invention) may all fall within the scope of the present invention.

[0171] The diagrams and / or illustrations contained in the accompanying drawings illustrate example implementations of one or more systems, methods and / or computer program products according to one or more embodiments of the present invention. It should be understood that one or more blocks in the accompanying drawings may represent components, fragments or portions of code, including one or more executable instructions for implementing specified logical functions. In some alternative implementations, the actions or functions identified in the blocks may occur in an order different from that shown in the accompanying drawings, or may occur simultaneously.

[0172] It should be understood that the blocks or steps shown in the drawings can be implemented by system components or computer program instructions. These instructions can be provided to the processor of any appropriate computer or other device, so that the instructions executed by the processor of the computer or other device can be established or generated to implement the functions or actions shown in the drawings.

Claims

1. A data management platform, comprising: a centralized database storing benchmark data, the centralized database being configured to receive one or more data sets from one or more data sources, wherein the centralized database allows one or more users to remotely access the centralized database; an analysis module, communicatively coupled to the centralized database, the analysis module comprising one or more analysis tools allowing data analysis of the one or more data sets; and A model module is communicatively coupled to the centralized database, the model module comprising one or more algorithms for executing a set of instructions on the one or more data sets, wherein the model module can be trained by the one or more algorithms by comparing the one or more data sets with the benchmark data to derive one or more output data sets having predictive information about the one or more data sets.

2. The data management platform according to claim 1, wherein: The one or more data sources are one or more data warehouses located in one or more countries.

3. The data management platform according to claim 1, wherein: The data management platform is configured to allow the one or more users to remotely access the centralized database via a network connection to transfer one or more data sets from one or more data sources to the centralized database, wherein the network connection allows the one or more users to analyze the one or more data sets using an analysis module including an analysis tool, and The network connection allows the one or more users to train a model module including one or more algorithms by comparing the one or more data sets with the benchmark data, thereby deriving an output data set having prediction information about the one or more data sets.

4. The data management platform according to claim 1, wherein: The analysis module including one or more analysis tools and the model module including one or more algorithms are dynamically updated and evolved, wherein each instance of the one or more analysis tools analyzes the one or more data sets, and each instance of the model module is trained by the one or more algorithms by comparing the one or more data sets with the benchmark data, thereby deriving one or more output data sets having predictive information about the one or more data sets.

5. The data management platform according to claim 1, wherein: The one or more analytical tools are selected from the following group: Tableau, Oracle Business Intelligence, IBM Cognos Analytics, SAS, Microsoft PowerBI, Amazon Redshift, Google BigQuery, Snowflake, Alteryx, Cloudera, Apache Hadoop, Google Vertex AI, Microsoft Azure Synapse, Microsoft Data Explorer, CosmosDB, Redis, Azure Cognitive Services, Azure Machine Learning, Spark, Databricks, Sqream, Confluent Kafka, Presto, Trino, Flare, HIDS and combinations thereof.

6. The data management platform according to claim 1, wherein: The one or more algorithms are selected from the following group: machine learning algorithms, artificial intelligence algorithms, deep learning algorithms, heuristic algorithms and combinations thereof.

7. The data management platform according to claim 1, wherein: The one or more data sets include customer usage information selected from the following group: customer voice call usage, customer screen time usage, customer data usage, customer messaging usage, customer device hardware specifications, customer transactions, customer interactions, customer behavior, customer revenue, network operations, network usage, network investment, internal operations, sales information, distribution information, agent operations, merchant operations, agent services, merchant operations, business-to-business products, business-to-business product services, digital products, OTT applications, customer value management, pricing, operations management, portfolio management, customer location, network location, network transmission, network configuration, network security and combinations thereof.

8. The data management platform according to claim 1, wherein: The predictive information includes information about customer behavior, thereby deriving user-specific product offers.

9. The data management platform according to claim 1, wherein: The centralized database is configured to receive only one or more data sets that have been anonymized and derive the anonymized data.

10. The data management platform according to claim 1, further comprising: a data transmission module, used for transmitting the model module to a database outside the centralized database, The model module including one or more algorithms can be applied to one or more data sets available on the database, thereby deriving an output data set having prediction information about the one or more data sets.

11. The data management platform according to claim 10, wherein: The model module can be stored on a storage medium.

12. The data management platform according to claim 11, wherein: The storage medium is selected from the group consisting of a non-transitory storage medium, a transient storage medium, and a combination thereof.

13. The data management platform according to claim 1, further comprising: An anonymization module located outside the centralized database, the anonymization module comprising: A data module, configured to receive the one or more data sets from one or more data sources; an anonymization algorithm, used to anonymize the one or more data sets to derive anonymized data; and A transmission module is used to transmit the anonymized data to the centralized database.

14. The data management platform according to claim 13, wherein: The anonymized data is used by the one or more algorithms to train the model module by comparing the anonymized data with benchmark data stored in the centralized database, thereby deriving an output data set having prediction information about the anonymized data.

15. The data management platform according to claim 14, wherein: The model module is configured to be trained using the anonymized data, and The model module is applied to one or more data sets located in a selected location to comply with country-specific data protection and privacy regulations.

16. The data management platform according to claim 1, further comprising: An interface module is communicatively coupled to the model module to receive prediction information about the one or more data sets, and the interface module is also communicatively coupled to one or more user devices to allow the user devices to access the prediction information about the one or more data sets.

17. A method for training a model using anonymized data and applying the model to one or more data sets at a selected location, the method comprising the steps of: A database storing benchmark data is provided, wherein: The database is capable of receiving data from one or more data sources; providing an external data source communicatively coupled to the database, wherein the one or more data sets stored by the external data source are anonymized by executing a set of instructions on the one or more data sets stored by the external data source, thereby deriving anonymized data; Using one or more analysis tools to allow data analysis of the data stored in the database; training a model comprising one or more algorithms by comparing the anonymized data to the benchmark data; The trained model is applied to one or more data sets located outside of the database to derive predictive information about the one or more data sets.

18. The method according to claim 17, further comprising the steps of: Storing the model on a storage medium allows the model to be applied to one or more data sets communicatively coupled to the storage medium.

19. A digital management system comprising: a computing device including a processor, the processor communicatively coupled to a memory capable of storing one or more data sets obtained from one or more data sources, the memory storing benchmark data; an analysis module including an analysis tool, the analysis module being communicatively coupled to the memory, wherein the processor is capable of executing a set of instructions to cause the analysis tool to perform an analysis operation on the one or more data sets; and A model module comprising one or more algorithms, the model module being communicatively coupled to the memory, wherein the model module is capable of being trained by the processor by applying the one or more algorithms by comparing the one or more data sets with benchmark data, thereby deriving an output data set having predictive information about the one or more data sets.

20. The digital management system according to claim 19, wherein: The memory is selected from the group consisting of a non-transitory storage medium, a transient storage medium, and combinations thereof.

Images