Artificial swarm intelligence via networked recommendation engines

By networking the basic recommendation engine with other recommendation engines, using the output and feedback data of multiple recommendation engines, the problem that existing recommendation engines cannot utilize cross-LOB and domain data is solved, improving the accuracy and quality of the recommendation system, and reducing the deviation.

CN112131458BActive Publication Date: 2024-11-12SAP PORTALS ISRAEL
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202010272708.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2019-06-25
Filing Date
2020-04-09
Publication Date
2024-11-12
Estimated Expiration
2040-04-09

AI Technical Summary

Technical Problem

Existing machine learning-based recommendation engines are unable to effectively utilize input data from other LOBs and domains, resulting in insufficient accuracy and quality of recommendations, and functional isolation leads to bias and isolation of recommendations.

Method used

By networking the basic recommendation engine with other recommendation engines, the output of multiple recommendation engines can be used to enhance the accuracy and quality of the recommendation system, and update the model through feedback data to reduce bias.

Benefits of technology

It improves the accuracy and quality of the recommendation system, reduces deviations, enhances the comprehensive utilization of recommendations in different fields, and improves the comprehensive effect of recommendations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112131458B_ABST
    Figure CN112131458B_ABST
Patent Text Reader

Abstract

Disclosed are technologies for implementing artificial swarm intelligence via networking of recommendation engines. In some example embodiments, a base recommendation engine performs a computer-implemented method, the method comprising: receiving a recommendation request; generating a first plurality of recommendations based on the recommendation request using a base recommendation model; obtaining one or more other recommendations from one or more other recommendation engines, each of the other recommendation engines being configured to generate other recommendations based on the recommendation request using a corresponding recommendation model; obtaining corresponding weights of (multiple) other recommendation engines; selecting at least one recommendation from the first plurality of recommendations and (multiple) other recommendations using (multiple) corresponding weights; and causing the selected (multiple) recommendations to be displayed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates generally to the technical field of electronic computer systems and, in various embodiments, to systems and methods for implementing artificial swarm intelligence via networking of recommendation engines. Background Art

[0002] Existing machine learning-based recommendation engines rely on input data that is considered directly relevant to the problem being solved. This data comes from sources within the same system, or from connected systems within the same Line of Business (LOB) or domain. Although directly relevant data from its own LOB helps to produce high-quality recommendations, this existing approach cannot identify useful input from other LOBs and domains because existing recommendation engines are not networked with other recommendation engines, nor do they leverage other recommendation engines. The technical problem of functional isolation of recommendation engines from each other leads to another technical problem of lack of accuracy and quality in the recommendations generated by the recommendation system. The present disclosure solves these and other technical problems that plague the computer functions of recommendation systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0003] Some example embodiments of the present disclosure are illustrated by way of example and not limitation in the accompanying figures in which like references indicate similar elements.

[0004] Figure 1 is a network diagram illustrating a client-server system according to some example embodiments.

[0005] Figure 2 is a block diagram illustrating enterprise applications and services in an enterprise application platform according to some example embodiments.

[0006] Figure 3 is a block diagram illustrating a recommendation system according to some example embodiments.

[0007] Figure 4 A network environment is shown in which a recommendation engine uses recommendations from other recommendation engines according to some example embodiments.

[0008] Figure 5 A network environment is shown in which a recommendation engine weights recommendations from other recommendation engines, according to some example embodiments.

[0009] Figure 6 Shown is a table of stored associations of corresponding weights of recommendations from other recommendation engines, according to some example embodiments.

[0010] Figure 7Another network environment is shown in which a recommendation engine uses recommendations from other recommendation engines via a central recommendation manager, according to some example embodiments.

[0011] Figure 8 Ranking of recommendations from a plurality of different recommendation engines for selecting a recommendation is shown according to some example embodiments.

[0012] Fig. 9 is a flow chart illustrating a method of implementing artificial swarm intelligence via networking of recommendation engines, according to some example embodiments.

[0013] Fig.10 is a block diagram of an example computer system on which the methods described herein may be performed, according to some example embodiments. DETAILED DESCRIPTION

[0014] Example methods and systems for implementing artificial swarm intelligence via networking of recommendation engines are disclosed. In the following description, for the purpose of explanation, numerous specific details are set forth in order to provide a thorough understanding of the example embodiments. However, it is apparent to those skilled in the art that the present embodiments can be implemented without these specific details.

[0015] The implementation of the features disclosed herein involves non-universal, non-traditional, and non-conventional operations or combinations of operations. By applying one or more solutions disclosed herein, some technical effects of the systems and methods disclosed herein are to provide a computer system configured to implement artificial swarm intelligence, wherein a base recommendation engine leverages recommendations from other recommendation engines via a networking system of recommendation engines. By networking with other recommendation engines, the base recommendation engine can identify useful recommendations that were previously unavailable, thereby improving the accuracy and quality of the basic recommendation system. As a result of the features disclosed herein, the operation of the recommendation system is improved. Other technical effects will also be apparent from this disclosure.

[0016] Even in the most advanced machine learning systems, one of the biggest challenges is the impact of biases, such as biased training data sets, on the recommendations of biased recommendation engines. Using the features of the present disclosure, this bias is reduced by networking different recommendation engines for their artificial swarm intelligence to identify further meaningful connections between input data to have a positive impact on improving recommendations, because the additional independent factors that positively affect the recommendations help to blur the bias. Instead of each recommendation engine working as an isolated silo, the system of the present disclosure networks different recommendation engines from the same and different LOBs and application domains. In addition, considering the outputs from a variety of other recommendation engines, including outputs from other LOBs or other application domains, enhances the recommendations of the base recommendation engine. In addition, by analyzing feedback data on the recommendations generated by these other recommendation engines, new connections can be identified and incorporated into the improved model of the base recommendation engine.

[0017] The methods or embodiments disclosed herein may be implemented as a computer system having one or more modules (e.g., hardware modules or software modules). These modules may be executed by one or more hardware processors of the computer system. In some example embodiments, a non-transitory machine-readable storage device may store an instruction set that, when executed by at least one processor, causes the at least one processor to perform the operations and method steps discussed within the present disclosure.

[0018] The details of one or more variations of the subject matter described herein are set forth in the accompanying drawings and the description that follows. Other features and benefits of the subject matter described herein will become apparent from the description and drawings, and from the claims.

[0019] Figure 1 1 is a network diagram illustrating a client-server system 100 according to some example embodiments. A platform (e.g., machine and software) in the example form of an enterprise application platform 112 provides server-side functionality to one or more clients via a network 114 (e.g., the Internet). For example, Figure 1 A client machine 116 with a programmatic client 118 (eg, a browser), a small device client machine 122 with a small device web client 120 (eg, a browser without a scripting engine), and a client / server machine 117 with a programmatic client 119 are shown.

[0020] Turning specifically to the example enterprise application platform 112, a network server 124 and an application programming interface (API) server 125 may be coupled to an application server 126 and provide network and programming interfaces to the application server 126. The application server 126 may in turn be coupled to one or more database servers 128, wherein the database server 128 facilitates access to one or more databases 130. The cross-functional service 132 may include a relational database module to provide support services for access to the database(s) 130 including a user interface library 136. The network server 124, the API server 125, the application server 126, and the database server 128 may host the cross-functional service 132. The application server 126 may also host a domain application 134.

[0021] The cross-functional services 132 provide services to users and processes that utilize the enterprise application platform 112. For example, the cross-functional services 132 may provide portal services (e.g., web services), database services, and connections to domain applications 134 for users operating the client machines 116, the client / server machines 117, and the small device client machines 122. In addition, the cross-functional services 132 may provide an environment for delivering enhancements to existing applications and for integrating third-party and legacy applications with existing cross-functional services 132 and domain applications 134. Furthermore, although Figure 1 The illustrated system 100 employs a client-server architecture, but embodiments of the present disclosure are certainly not limited to such an architecture and may equally find application in distributed or peer-to-peer architecture systems.

[0022] The enterprise application platform 112 can improve (e.g., increase) the accessibility of data across different environments of a computer system architecture. For example, when testing an instance of a software solution in a development environment, the enterprise application platform 112 can effectively and efficiently enable users to use real data created by one or more end users of a deployed instance of the software solution in a production environment. Figure 2-Figure 8 Enterprise application platform 112 is described in more detail.

[0023] Figure 2 1 is a block diagram illustrating enterprise applications and services in an enterprise application platform 112 according to an example embodiment. The enterprise application platform 112 may include cross-functional services 132 and domain applications 134. The cross-functional services 132 may include a portal module 140, a relational database module 142, a connector and messaging module 144, an API module 146, and a development module 148.

[0024] The portal module 140 can enable a single point of access to other cross-functional services 132 and domain applications 134 for the client machines 116, small device client machines 122, and client / server machines 117. The portal module 140 can be used to process, author, and maintain web pages that present content (e.g., user interface elements and navigation controls) to users. In addition, the portal module 140 can enable user roles, which is a construct that associates a role with a dedicated environment that a user uses to perform tasks, utilize services, and exchange information with other users within a defined scope. For example, a role can determine the content available to a user and the activities that the user can perform. The portal module 140 includes a generation module, a communication module, a reception module, and a reproduction module. In addition, the portal module 140 can comply with network service standards and / or utilize various Internet technologies, including Advanced business application programming language for J2EE and SAP and Web Dynpro, XML, JCA, JAAS, X.509, LDAP, WSDL, WSRR, SOAP, UDDI, and

[0025] The relational database module 142 can provide support services for accessing the (multiple) databases 130 including the user interface library 136. The relational database module 142 can provide support for object-relational mapping, database independence, and distributed computing. The relational database module 142 can be used to add, delete, update, and manage database elements. In addition, the relational database module 142 can comply with database standards and / or utilize various database technologies, including SQL, SQLDBC, Oracle, MySQL, Unicode, JDBC, etc.

[0026] The connector and messaging module 144 can enable communication across different types of messaging systems utilized by the cross-functional services 132 and the domain applications 134 by providing a common messaging application processing interface. The connector and messaging module 144 can enable asynchronous communication on the enterprise application platform 112.

[0027] The API module 146 can enable the development of service-based applications by exposing interfaces to existing and new applications as services.A repository can be included in the platform as a central location to find available services when building applications.

[0028] The development module 148 may provide a development environment for adding, integrating, updating, and extending software components on the enterprise application platform 112 without affecting existing cross-functional services 132 and domain applications 134 .

[0029] Turning to the domain application 134, the customer relationship management application 150 can enable access to multiple data sources and business processes and can facilitate the collection and storage of relevant personalized information from the multiple data sources and business processes. Enterprise personnel responsible for developing buyers into long-term customers can use the customer relationship management application 150 to provide assistance to buyers throughout the customer engagement cycle.

[0030] Enterprise personnel can utilize financial applications 152 and business processes to track and control financial transactions within the enterprise application platform 112. Financial applications 152 can facilitate the execution of operational, analytical, and collaborative tasks associated with financial management. Specifically, financial applications 152 can enable the execution of tasks related to financial responsibility, planning, forecasting, and managing financial costs.

[0031] Enterprise personnel and business processes can manage, deploy, and track enterprise personnel using human resource applications 154. Specifically, human resource applications 154 can enable analysis of human resource issues and facilitate human resource decisions based on real-time information.

[0032] The product lifecycle management application 156 can enable the management of the product throughout its lifecycle. For example, the product lifecycle management application 156 can enable collaborative engineering, custom product development, project management, asset management, and quality management between business partners.

[0033] Supply chain management applications 158 can enable monitoring of performance observed in the supply chain. Supply chain management applications 158 can facilitate adherence to production schedules and on-time delivery of products and services.

[0034] Third-party applications 160 as well as legacy applications 162 can be integrated with domain applications 134 and leverage cross-functional services 132 on the enterprise application platform 112 .

[0035] Figure 3 300 according to some example embodiments. In some embodiments, the recommendation system 300 includes any combination of one or more of a recommendation module 310, a networking module 320, a machine learning module 330, and one or more databases 340. The modules 310, 320, and 330 and the (multiple) databases 340 may reside on a computer system or other machine having a memory and at least one processor (not shown). In some embodiments, the modules 310, 320, and 330 are incorporated into Figure 1 The application server(s) 126 in the embodiment of the present invention and the database(s) 340 are merged into Figure 1 However, other configurations of modules 310, 320, and 330 and database(s) 340 are contemplated to be within the scope of the present disclosure.

[0036] In some example embodiments, one or more of modules 310, 320, and 330 are configured to provide various user interface functions, such as generating a user interface, interactively presenting a user interface to a user, receiving information from a user (e.g., interaction with a user interface), etc. Presenting information to a user may include causing information to be presented to a user (e.g., using instructions to convey information to a device to present information to a user). Information may be presented using various means, including visually displaying information and using other device outputs (e.g., audio, tactile, etc.). Similarly, information may be received via various means, including alphanumeric input or other device input (e.g., one or more touch screens, cameras, tactile sensors, light sensors, infrared sensors, biometric sensors, microphones, gyroscopes, accelerometers, other sensors, etc.). In some example embodiments, one or more of modules 310, 320, and 330 are configured to receive user input. For example, one or more of modules 310, 320, and 330 may present one or more GUI elements (e.g., drop-down menus, selectable buttons, text fields) with which a user may submit input. In some example embodiments, one or more of modules 310, 320, and 330 are configured to perform various communication functions to facilitate the functionality described herein, such as communicating with a computing device (e.g., client machine 122, client machine 116, or client machine 117) via network 114 using a wired or wireless connection.

[0037] In some example embodiments, the recommendation system 300 is configured to enable a base recommendation engine to be networked with other recommendation engines to use the recommendations of the other recommendation engines when responding to recommendation requests. Figure 4 4 shows a network environment 400 in which a recommendation engine 410 uses recommendations 420 from other recommendation engines 410, according to some example embodiments. Figure 4 The example shown includes three different recommendation engines 410 - 1 , 410 - 2 , and 410 - 3 , but other numbers of recommendation engines are within the scope of the present disclosure. In some example embodiments, each recommendation engine 410 includes its own dedicated instance of the recommendation system 300 into which the recommendation engine 410 is incorporated.

[0038] In any particular instance where a recommendation engine 410 receives a recommendation request, the recommendation engine 410 that receives the recommendation request is referred to as a base recommendation engine, and the other recommendation engines 410 from which the base recommendation engine obtains other recommendations are referred to as other recommendation engines. In one example, recommendation engine 410-1 receives the recommendation request and is therefore the base recommendation engine for this example, while other recommendation engines 410-2 and 410-3 are other recommendation engines.

[0039] Each of recommendation engines 410-1, 410-2, and 410-3 can be configured to serve specific categories of organizational functions that are different from each other (such as different LOBs). In some example embodiments, recommendation engines 410-1, 410-2, and 410-3 each have their own application domain that is different from the application domains of other recommendation engines 410. An application domain is a mechanism (e.g., similar to a process in an operating system) for isolating executing software applications from each other so that they do not affect each other. Each application domain has its own virtual address space, which is used to define the resource scope of the application domain.

[0040] In one example, recommendation engine 410-1 is configured to serve the manufacturing department or department of an organization via its own corresponding application domain, recommendation engine 410-2 is configured to serve the marketing department or department of an organization via its own corresponding application domain, and recommendation engine 410-3 is configured to serve the legal department or department of an organization via its own corresponding application domain. Each recommendation engine 410 is configured to generate recommendations 420 that can be used by other recommendation engines 410. For example, base recommendation engine 410-1 is configured to generate recommendation 420-1, other recommendation engine 410-2 is configured to generate recommendation 420-2, and other recommendation engine 410-3 is configured to generate recommendation 420-3.

[0041] In some example embodiments, the recommendation module 310 of the basic recommendation engine 410-1 is configured to receive a recommendation request from a computing device. The recommendation request may include a request for any type of recommendation. For example, the recommendation request may include a request for online content or a request for a recommendation of an action to be performed. Examples of online content include, but are not limited to: search results of a search query included in the recommendation request, online coupons displayed to an end user, online advertisements displayed to an end user (e.g., online banner ads), recommendations of online media content (e.g., movie recommendations), online feed items (e.g., news feed content items), and online job postings. Examples of recommendations of actions to be performed include, but are not limited to, recommendations to move inventory items from one location to another, recommendations to manufacture a specific quantity of a specific product, and recommendations to purchase specific equipment items for manufacturing.

[0042] In some example embodiments, the recommendation module 310 of the basic recommendation engine 410-1 is configured to generate a first plurality of recommendations based on the recommendation request using a basic recommendation model. The basic recommendation model may include functionality configured to generate a recommendation score for a recommendation candidate based on one or more factors or parameters, and the basic recommendation model may then select one or more of the recommendation candidates for use as a recommendation based on the recommendation score. For example, the basic recommendation model may rank the recommendation candidates based on their recommendation scores (e.g., in descending order), and then select a top-ranked portion of the recommendation candidates (e.g., the top five ranked recommendation candidates, the top five percent of the recommendation candidates), or a threshold recommendation score may be used to select a recommendation candidate having a corresponding recommendation score that satisfies the threshold recommendation score (e.g., a recommendation candidate having a recommendation score of 0.75 or higher).

[0043] In some example embodiments, the networking module 320 of the base recommendation engine 410-1 is configured to obtain a corresponding set of one or more other recommendations 420 from one or more other recommendation engines 410. Each of the one or more other recommendation engines 410 may be configured to generate a corresponding set of one or more other recommendations 420 based on the recommendation request using a corresponding recommendation model that is different from the base recommendation model used by the base recommendation engine 410. For example, although the corresponding recommendation models of the other recommendation engines 410-2 and 420-3 may also include functionality configured to generate a recommendation score for a recommendation candidate based on one or more factors or parameters, and may be configured to select one or more of the recommendation candidates for use as a recommendation based on the recommendation score, the functionality used by the other recommendation engine 410-2 may be different from the functionality used by the base recommendation engine 410-1 and the other recommendation engines 410-3, and the functionality used by the other recommendation engine 410-3 may also be different from the functionality used by the base recommendation engine 410-1 and the other recommendation engines 410-3.

[0044] In some example embodiments, the networking module 320 of the base recommendation engine 410-1 is configured to, for each of the other recommendation engines 410-2 and 410-3, obtain the corresponding weights of the other recommendation engines 410-2 and 410-3 for weighting the recommendations 420-2 and 420-3 from the other recommendation engines 410-2 and 410-3. Figure 5 5 shows a network environment 500 in which a recommendation engine weights recommendations from other recommendation engines, according to some example embodiments. Figure 5 In the network environment 500, it is similar to Figure 4 400, except that the recommendations 420 are replaced by their corresponding weights. Figure 5As shown by the arrows and weights in , the recommendation system 300 implements independent bidirectional weights for recommendations between different recommendation engines 410. These weights represent both the reliability and credibility of the recommendations. Figure 5 In the example shown, recommendation 420-1 generated by recommendation engine 410-1 has a weight of 0.4 when used by recommendation engine 410-2 and a weight of 0.7 when used by recommendation engine 410-3, recommendation 420-2 generated by recommendation engine 410-2 has a weight of 0.4 when used by recommendation engine 410-1 and a weight of 0.5 when used by recommendation engine 410-3, and recommendation 420-3 generated by recommendation engine 410-3 has a weight of 0.3 when used by recommendation engine 410-1 and a weight of 0.7 when used by recommendation engine 410-2.

[0045] Figure 6 A table 600 of stored associations of corresponding weights of recommendations from other recommendation engines is shown according to some example embodiments. Each table 600 corresponds to a weight used by a base recommendation engine for recommendations from other recommendation engines. For example, table 600-1 includes corresponding weights used by recommendation engine 410-1 for recommendations obtained from recommendation engines 410-2 and 410-3, table 600-2 includes corresponding weights used by recommendation engine 410-2 for recommendations obtained from recommendation engines 410-1 and 410-3, and table 600-3 includes corresponding weights used by recommendation engine 410-3 for recommendations obtained from recommendation engines 410-1 and 410-2. The weights may be stored, managed, and retrieved from database(s) 340.

[0046] In some example embodiments, the recommendation module 310 of the base recommendation engine 410-1 is configured to select at least one recommendation from the pool of recommendations formed by the first plurality of recommendations 420-1 generated by the base recommendation engine 410-1 and each corresponding set of one or more other recommendations 420-1 and 420-3 from one or more other recommendation engines 410-2 and 420-3, using each corresponding weight of each set of one or more other recommendations 420-1 and 420-3 from one or more other recommendation engines 410-2 and 420-3. The recommendation module 310 may select (multiple) recommendations based on recommendation scores. For example, the recommendation module 310 may rank the recommendation pool based on their recommendation scores (e.g., in descending order), and then select a top-ranked portion of the recommendations (e.g., the top five recommendations, the top five percent recommendations), or may use a threshold recommendation score to select a recommendation with a corresponding recommendation score that satisfies the threshold recommendation score (e.g., a recommendation with a recommendation score of 0.75 or higher).

[0047] In some example embodiments, the recommendation module 310 of the base recommendation engine 410-1 is configured to cause the selected at least one recommendation to be displayed on the computing device. The selected (multiple) recommendation(s) may be displayed in association with the recommendation request. For example, the selected (multiple) recommendation(s) may include search results displayed in association with the search query of the recommendation request. In some example embodiments, the selected (multiple) recommendation(s) include at least one recommendation for online content or at least one recommendation for performing an action. However, other types of recommendations are also within the scope of the present disclosure.

[0048] In some example embodiments, the machine learning module 330 of the basic recommendation engine 410-1 is configured to receive corresponding feedback data for each of the selected (multiple) recommendations, and then use the received feedback data as training data in at least one machine learning operation to modify the corresponding weight of at least one of the one or more other recommendation engines stored in the database. The feedback data may indicate the reaction of the user of the computing device to the display of the selected (multiple) recommendations on the computing device. For example, the basic recommendation engine 410-1 and the other recommendation engines 410-2 and 410-3 may track whether the user chooses to perform an online action (e.g., view, purchase) for the (multiple) recommendations or ignore or reject the (multiple) recommendations. These tracked reactions can be stored in the (multiple) databases 340 and used as feedback data. The machine learning module 330 can aggregate and combine different recommendation algorithm scores, feedback data, and business priorities, and update the recommendation model (e.g., the weights used in the recommendation model) based on the end user feedback, so that the recommendation system 300 can continuously learn and adjust. Over time, the effectiveness of recommendations from one recommendation engine 410 against the recommendations of another recommendation engine 410 will increase or decrease dynamic weight accordingly. The success of recommendations can be evaluated via methods such as click-through rate and purchase / shopping cart abandonment rate. As a result, each recommendation engine can better understand how useful each other recommendation engine is in enhancing its own recommendations.

[0049] although Figure 4 The base recommendation engine 410-1 is shown to obtain other recommendations 420-2 and 420-3 directly from other recommendation engines 410-2 and 420-3 via a direct network connection between the base recommendation engine 410-1 and the other recommendation engines 410-2 and 410-3, but in some example embodiments, the other recommendations 420-2 and 420-3 are obtained via a management server that is separate and different from the base recommendation engine 410-1. Figure 7Another network environment 700 is shown in which a recommendation engine 410 uses recommendations 420 from other recommendation engines 410 via a central recommendation manager 710, according to some example embodiments. In some example embodiments, the central management server 710 is configured to obtain a corresponding set of one or more other recommendations 420-2 and 420-3 from each of the other recommendation engines 410-2 and 420-3, and then send the corresponding set of one or more other recommendations 420-2 and 420-3 to the base recommendation engine 410-1. The weights of the recommendation engines 410 can be stored, managed, and obtained from the central recommendation manager 710.

[0050] In some example embodiments, the base recommendation engine 410-1 is configured to detect a switching event corresponding to one or more of the other recommendation engines 410-2 and 420-3 (such as a network disconnection or other unavailability of the other recommendation engines 420-2 and 420-3), and based on the detected switching event, obtain a corresponding set of one or more other recommendations from the other recommendation engines corresponding to the switching event via the central management server 710. As a result of this switching capability, the recommendation system 300 solves the technical problems of network disconnection and other situations that make another recommendation engine unavailable.

[0051] Figure 8 A ranking 800 of recommendations from multiple different recommendation engines is shown, which is used to select recommendations, according to some example embodiments. In some example embodiments, the recommendation module 310 sorts the recommendations generated by the base recommendation engine 410-1 and the recommendations generated by and obtained from the other recommendation engines 410-2 and 410-3 (such as in descending order of recommendation scores) based on the corresponding recommendation scores of the recommendations. The recommendation module 310 can then select one or more of the recommendations based on the ranking 800 to respond to the recommendation request. For example, the recommendation module 310 can select the top five recommendations or the top five percent of recommendations.

[0052] Fig. 9 900 is a flow chart illustrating a method 900 for implementing artificial swarm intelligence via networking of recommendation engines according to some example embodiments. The method 900 may be performed by processing logic that may include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions running on a processing device), or a combination thereof. In an example embodiment, one or more operations of the method 900 are performed by a base recommendation engine, wherein Figure 3 The recommendation system or any combination of one or more of its modules 310, 320 and 330 is implemented as described above. However, in some example embodiments, one or more operations of method 900 are additionally or alternatively performed by a management server (e.g., Figure 5 The recommendation manager 510 in is executed.

[0053] At operation 910, the base recommendation engine receives a recommendation request from a computing device. For example, the base recommendation engine may receive a request for online content (such as a search query), or a request for a recommendation for an action to be performed (such as a request for a recommendation for an action that an organization or organization member should perform).

[0054] At operation 920, the base recommendation engine generates a first plurality of recommendations based on the recommendation request using a base recommendation model. The base recommendation model may take into account one or more parameters included in or otherwise associated with the recommendation request, such as search terms of a search query or an identification of a user to whom the recommendation request is being submitted. Other types of parameters are also within the scope of the present disclosure.

[0055] At operation 930, the base recommendation engine obtains a corresponding set of one or more other recommendations from one or more other recommendation engines. In some example embodiments, each of the one or more other recommendation engines is configured to generate a corresponding set of one or more other recommendations based on a recommendation request using a corresponding recommendation model that is different from the base recommendation model. The base recommendation engine and each of the one or more other recommendation engines may form a group of recommendation engines, each recommendation engine being configured to serve a specific category of organizational functions that is different from the specific category of organizational functions of all other recommendation engines in the group. The base recommendation engine and each of the one or more other recommendation engines may form a group of recommendation engines, each recommendation engine having their own application domain that is different from the application domains of all other recommendation engines in the group. Other differences between different recommendation engines are also within the scope of the present disclosure.

[0056] In some example embodiments, for each of the one or more other recommendation engines, a corresponding set of one or more other recommendations is obtained from the one or more other recommendation engines via a direct network connection between the base recommendation engine and the one or more other recommendation engines. In other example embodiments, for each of the one or more other recommendation engines, a corresponding set of one or more other recommendations is obtained from the one or more other recommendation engines via a management server that is separate and different from the base recommendation engine, wherein the management server is configured to obtain the corresponding set of one or more other recommendations from each of the one or more other recommendation engines, and send the corresponding set of one or more other recommendations obtained from each of the one or more other recommendation engines to the base recommendation engine.

[0057] In some example embodiments, the base recommendation engine is configured to obtain one or more corresponding sets of other recommendations from one or more other recommendation engines via a direct network connection between the base recommendation engine and the one or more other recommendation engines, detect a switching event corresponding to one of the one or more other recommendation engines, and based on the detected switching event, obtain one or more corresponding sets of other recommendations from the other recommendation engines corresponding to the switching event via a management server separate and different from the base recommendation engine. The management server may be configured to obtain one or more corresponding sets of other recommendations from the other recommendation engines corresponding to the switching event, and then send the one or more corresponding sets of other recommendations obtained from the other recommendation engines corresponding to the switching event to the base recommendation engine.

[0058] At operation 940, for each of the one or more other recommendation engines, the base recommendation engine obtains the corresponding weight of the other recommendation engines from the database. In some example embodiments, the base recommendation engine selects one or more other recommendation engines from the plurality of recommendation engines to be used to obtain the corresponding set of one or more other recommendations based on the corresponding weight of each of the selected one or more other recommendation engines. In some example embodiments, selecting the one or more other recommendation engines includes omitting at least one of the plurality of recommendation engines from the selection based on the corresponding weight of the at least one recommendation engine in the plurality of recommendation engines.

[0059] Thereafter, the base recommendation engine selects at least one recommendation from the first plurality of recommendations and each corresponding set of one or more other recommendations from the one or more other recommendation engines using each corresponding weight of each set of one or more other recommendations from the one or more other recommendation engines at operation 950. In some example embodiments, the selected at least one recommendation includes at least one recommendation for online content or at least one recommendation for performing an action.

[0060] At operation 960, the base recommendation engine causes the selected at least one recommendation to be displayed on the computing device. In some example embodiments, the selected at least one recommendation is displayed in association with the recommendation request. For example, the selected at least one recommendation may include search results displayed in association with the search query of the recommendation request. In some example embodiments, the selected at least one recommendation includes at least one recommendation for online content or at least one recommendation to perform an action. However, other types of recommendations are also within the scope of the present disclosure.

[0061] At operation 970, the base recommendation engine 410 receives corresponding feedback data for each of the selected at least one recommendation. In some example embodiments, the feedback data indicates a reaction of a user of the computing device to the display of the selected at least one recommendation on the computing device. For example, the base recommendation engine and other recommendation engines may track whether the user chooses to perform an online action (e.g., view, purchase) for the recommendation(s) or to ignore or reject the recommendation(s).

[0062] As shown in operation 980, the base recommendation engine 410 uses the received feedback data as training data in at least one machine learning operation to modify the corresponding weight of at least one of the one or more other recommendation engines stored in the database. The method 900 can then return to operation 910, where the base recommendation engine receives another recommendation request, and the remaining operations of the method 900 are performed using the modified weight(s).

[0063] It is contemplated that any other features described in this disclosure may be incorporated into method 900 .

[0064] Features disclosed herein provide several technical benefits to the computer functions of the recommendation system 300. New connections are also identified, which allows new input data to directly enhance recommendation engines that it has not been fed to before, which can be represented by the situation where the weight on the arrow pointing from the new recommendation engine in the network is above a threshold. For example, it may not have been known in the past that an environmental sustainability recommendation engine would enhance the results of a marketing recommendation engine. However, such a connection can be revealed via the networked system of the present disclosure, which determines that in fact the purchase decision is also affected by the "green factor" of the product. By merging recommendations from other recommendation engines in the network, the results of each recommendation engine on the networked system are enhanced. For example, such a network can establish how effective a legal recommendation engine and a marketing recommendation engine are as inputs to a manufacturing recommendation engine. As a result, the recommendations of the manufacturing recommendation engine will be more effective.

[0065] In practice, new context-based recommendation engines are created dynamically. For example, since legal recommendations and marketing recommendations may be meaningful inputs to a manufacturing recommendation engine, we effectively have a new recommendation engine. It will certainly contain the manufacturing recommendation engine, but it will also include a legal recommendation engine with weight "a" and a marketing recommendation engine with weight "b". If legal's recommendations are more effective for manufacturing than marketing's recommendations, then "a" will be greater than "b". Furthermore, since these weights are directional, the weight that the manufacturing recommendation engine has over the legal recommendation engine may be completely different than the weight it has in the opposite direction. Furthermore, these weights are dynamic, and over time the weight of one recommendation engine over another changes as its recommendations become more relevant to that other recommendation engine.

[0066] Certain embodiments are described herein as including logic or multiple components, modules, or mechanisms. A module may constitute a software module (e.g., code contained on a machine-readable medium or in a transmission signal) or a hardware module. A hardware module is a tangible unit capable of performing a specific operation and may be configured or arranged in a particular manner. In an example embodiment, one or more computer systems (e.g., stand-alone, client, or server computer systems) or one or more hardware modules of a computer system (e.g., a processor or a group of processors) may be configured as a hardware module by software (e.g., an application or application portion), wherein the hardware module operates to perform specific operations described herein.

[0067] In various embodiments, hardware modules may be implemented mechanically or electronically. For example, a hardware module may include dedicated circuitry or logic that is permanently configured (e.g., as a dedicated processor (such as a field programmable gate array (FPGA) or an application specific integrated circuit (ASIC))) to perform specific operations. A hardware module may also include programmable logic or circuitry (e.g., contained within a general purpose processor or other programmable processor) that is temporarily configured by software to perform specific operations. It should be understood that the decision to implement a hardware module mechanically, in a dedicated and permanently configured circuit, or in a temporarily configured circuit (e.g., configured by software) may be driven by cost and time considerations.

[0068] Thus, the term "hardware module" should be understood to include a tangible entity that is physically constructed, permanently configured (e.g., hardwired), or temporarily configured (e.g., programmed) to operate in a particular manner and / or perform specific operations described herein. Considering embodiments where hardware modules are temporarily configured (e.g., programmed), each hardware module need not be configured or instantiated at any one time. For example, where the hardware modules include a general-purpose processor configured using software, the general-purpose processor can be configured as corresponding different hardware modules at different times. The software can configure the processor accordingly, for example, to constitute a specific hardware module at one time, and to constitute different hardware modules at different times.

[0069] Hardware modules can provide information to other hardware modules and can also receive information from other hardware modules. Therefore, the described hardware modules can be regarded as communicatively coupled. In the case where multiple such hardware modules exist at the same time, communication can be achieved by signal transmission (for example, by connecting appropriate circuits and buses of the hardware modules). In embodiments where multiple hardware modules are configured or instantiated at different times, communication between these hardware modules can be achieved, for example, by storing and retrieving information in a memory structure accessible to multiple hardware modules. For example, a hardware module can perform an operation and store the output of the operation in a memory device to which it is communicatively coupled. Then, another hardware module can access the storage device at a later time to retrieve and process the stored output. The hardware module can also initiate communication with an input or output device and can operate on a resource (for example, a collection of information).

[0070] The various operations of the example methods described herein may be performed at least in part by one or more processors that are temporarily configured (e.g., by software) or permanently configured to perform the related operations. Whether temporarily or permanently configured, these processors may constitute processor-implemented modules, wherein the processor-implemented modules operate to perform one or more operations or functions. In some example embodiments, the modules mentioned herein may include processor-implemented modules.

[0071] Similarly, the methods described herein may be implemented by a processor at least in part. For example, at least some of the operations of the method may be performed by one or more processors or a module implemented by the processor. The execution of certain operations may be distributed among one or more processors, not only residing in a single machine, but also deployed across multiple machines. In some example embodiments, one or more processors may be located in a single location (e.g., in a home environment, an office environment, or as a server farm), while in other embodiments, the processor may be distributed across multiple locations.

[0072] One or more processors may also support execution of related operations in a "cloud computing" environment or as "software as a service" (SaaS). For example, at least some operations may be performed by a group of computers (as an example of a machine including a processor), and the operations may be performed via a network (e.g., Figure 1 network 114) and accessed via one or more appropriate interfaces (e.g., APIs).

[0073] Example embodiments may be implemented in digital electronic circuitry, or in computer hardware, firmware, software, or a combination of these. Example embodiments may be implemented using a computer program product, for example, a computer program tangibly embodied in an information carrier (e.g., in a machine-readable medium) for execution by a data processing apparatus (e.g., a programmable processor, a computer, or multiple computers), or to control the operation of the data processing apparatus.

[0074] A computer program may be written in any form of programming language, including compiled or interpreted languages, and it may be deployed in any form, including as a stand-alone program or as a module, subroutine, or other unit suitable for use in a computing environment. A computer program may be deployed to be executed on one computer or on multiple computers at one site or distributed across multiple sites and interconnected by a communication network.

[0075] In example embodiments, operations may be performed by one or more programmable processors executing a computer program to perform functions by operating on input data and generating output. Method operations may also be performed by special purpose logic circuits (e.g., FPGAs or ASICs), and the apparatus of example embodiments may be implemented as special purpose logic circuits.

[0076] The computing system may include a client and a server. The client and the server are usually far away from each other and usually interact through a communication network. The relationship between the client and the server occurs by means of a computer program that runs on each computer and has a client-server relationship between each other. In the embodiment of the deployment of the programmable computing system, it should be understood that both hardware and software architectures are worthy of consideration. Specifically, it should be understood that it can be a design choice to select whether to implement certain functions in permanently configured hardware (e.g., ASIC), or in temporarily configured hardware (e.g., a combination of software and a programmable processor), or in a combination of permanently and temporarily configured hardware. The hardware (e.g., machine) and software architecture that can be deployed in various example embodiments are listed below.

[0077] Fig.101024. It is a block diagram of a machine in the example form of a computer system 1000, in which instructions 1024 for causing the machine to perform any one or more of the methods discussed herein can be executed. In alternative embodiments, the machine operates as a standalone device, or can be connected (e.g., networked) to other machines. In a networked deployment, the machine can operate in the capacity of a server or client machine in a server-client network environment, or as a peer machine in a peer-to-peer (or distributed) network environment. The machine can be a personal computer (PC), a tablet PC, a set-top box (STB), a personal digital assistant (PDA), a cellular phone, a network device, a network router, a switch or a bridge, or any machine capable of executing instructions (sequential or otherwise) specifying the actions to be taken by the machine. In addition, although only a single machine is shown, the term "machine" should also be understood to include any collection of machines that execute one or more instruction sets to perform any one or more of the methods discussed herein, either individually or in combination.

[0078] The example computer system 1000 includes a processor 1002 (e.g., a central processing unit (CPU), a graphics processing unit (GPU), or both), a main memory 1004, and a static memory 1006, which communicate with each other via a bus 1008. The computer system 1000 may also include a graphics or video display unit 1010 (e.g., a liquid crystal display (LCD) or a cathode ray tube (CRT)). The computer system 1000 also includes an alphanumeric input device 1012 (e.g., a keyboard), a user interface (UI) navigation (or cursor control) device 1014 (e.g., a mouse), a storage unit (e.g., a disk drive unit) 1016, an audio or signal generating device 1018 (e.g., a speaker), and a network interface device 1020.

[0079] The storage unit 1016 includes a machine-readable medium 1022 on which are stored one or more sets of data structures and instructions 1024 (e.g., software) that embody or are used by any one or more of the methodologies or functionality described herein. During execution by the computer system 1000, the instructions 1024 may also reside, completely or at least partially, in the main memory 1004 and / or the processor 1002, which also constitute machine-readable media. The instructions 1024 may also reside, completely or at least partially, in the static memory 1006.

[0080] Although the machine-readable medium 1022 is shown as a single medium in the example embodiment, the term "machine-readable medium" may include a single medium or multiple media (e.g., a centralized or distributed database, and / or associated caches and servers) storing one or more instructions 1024 or data structures. The term "machine-readable medium" should also be understood to include any tangible medium that can store, encode, or carry instructions executed by a machine and cause the machine to perform any one or more of the methods of the present embodiment, or any tangible medium that can store, encode, or carry data structures utilized by or associated with these instructions. The term "machine-readable medium" should be understood accordingly to include, but is not limited to, solid-state memory and optical and magnetic media. Specific examples of machine-readable media include non-volatile memory, including semiconductor storage devices (e.g., erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), and flash memory devices as examples); magnetic disks, such as internal hard disks and removable disks; magneto-optical disks; and compact disk read-only memory (CD-ROM) and digital versatile disk (or digital video disk) read-only memory (DVD-ROM) optical disks.

[0081] Instructions 1024 may also be sent or received over a communication network 1026 using a transmission medium. Instructions 1024 may be transmitted using a network interface device 1020 and any of a number of well-known transmission protocols (e.g., HTTP). Examples of communication networks include LANs, WANs, the Internet, mobile phone networks, POTS networks, and wireless data networks (e.g., WiFi and WiMax networks). The term "transmission medium" shall be understood to include any intangible medium capable of storing, encoding, or carrying instructions for execution by a machine, and includes digital or analog communication signals or other intangible media that facilitate communication of such software.

[0082] Each feature and teaching disclosed herein may be utilized alone or in combination with other features and teachings to provide systems and methods for blind spot implementation in neural networks. Representative examples utilizing many of these additional features and teachings (both individually and in combination) are also described in detail with reference to the accompanying drawings. This detailed description is intended only to teach those skilled in the art further details of practicing certain aspects of the present teachings and is not intended to limit the scope of the claims. Therefore, the combination of features disclosed above in the detailed description may not be necessary to practice the broadest teachings, but rather is taught to describe only specific representative examples of the present teachings.

[0083] Some parts of the detailed description of this paper are presented in terms of the symbolic representation of the operation of the data bit in the computer memory. These algorithmic descriptions and representations are used by the technical personnel in the field of data processing to convey the means of the essence of their work to other technical personnel in the field most effectively. Here, and generally, algorithms are generally understood to be self-consistent step sequences (self-consistent of steps) that realize the desired result. These steps are steps that require physical manipulation of physical quantities. Usually, although not necessarily, these quantities take the form of electrical or magnetic signals that can be stored, transferred, combined, compared, and otherwise manipulated. It has been proven that it is sometimes convenient to refer to these signals as bits, values, elements, symbols, characters, terms, numbers, etc., mainly for the reasons of public use.

[0084] It should be remembered, however, that these and similar terms are all to be associated with the appropriate physical quantities and are merely convenient labels applied to these quantities. Unless otherwise stated, it will be apparent from the following discussion that throughout the description, discussions utilizing terms such as "processing" or "computing" or "calculating" or "determining" or "displaying" refer to the actions and processes of a computer system or similar electronic computing device that manipulates and transforms data represented as physical (electronic) quantities within the computer system's registers and memories into other data similarly represented as physical quantities within the computer system's memories or registers or other such information storage, transmission or display devices.

[0085] The present disclosure also relates to an apparatus for performing the operations herein. The apparatus may be specially constructed for the desired purpose, or it may include a general-purpose computer selectively activated or reconfigured by a computer program stored in the computer. Such a computer program may be stored in a computer-readable storage medium, such as, but not limited to, any type of disk (including floppy disks, optical disks, CD-ROMs, and magneto-optical disks), read-only memory (ROM), random access memory (RAM), EPROM, EEPROM, magnetic or optical cards, or any type of medium suitable for storing electronic instructions, and each is coupled to a computer system bus.

[0086] The example methods or algorithms presented herein are not inherently related to any specific computer or other device. Various general purpose systems, computer servers, or personal computers can be used together with the program according to the teachings herein, or it can be proved that it is convenient to build a more dedicated device to perform the method steps disclosed herein. The structure of various these systems will appear from the description of this paper. It will be understood that the teachings of the disclosure described herein can be implemented using a variety of programming languages.

[0087] Furthermore, various features of the representative examples and dependent claims may be combined in ways not specifically and explicitly enumerated in order to provide additional useful embodiments of the present teaching. It is also clearly noted that all value ranges or indications of groups of entities disclose every possible intermediate value or intermediate entity for the purpose of original disclosure, as well as for the purpose of limiting the claimed subject matter. It should also be clearly noted that the sizes and shapes of the components shown in the figures are designed to help understand how to practice the present teaching, but are not intended to limit the sizes and shapes shown in the examples.

[0088] Although the embodiments have been described with reference to specific example embodiments, it is apparent that various modifications and changes may be made to these embodiments without departing from the broader spirit and scope of the present disclosure. Therefore, the specification and the drawings are to be considered illustrative rather than restrictive. The drawings constituting a part of this document illustrate specific embodiments in which the present subject matter may be practiced by way of illustration and not limitation. The illustrated embodiments are described in sufficient detail to enable those skilled in the art to practice the teachings disclosed herein. Other embodiments may be utilized and derived therefrom so that structural and logical substitutions and changes may be made without departing from the scope of the present disclosure. Therefore, this detailed description should not be understood in a restrictive sense, and the scope of the various embodiments is limited only by the appended claims and the full scope of equivalents to which these claims are assigned.

[0089] These embodiments of the subject matter of the present invention may be referred to herein individually and / or collectively as the term "invention", which is merely for convenience and is not intended to actively limit the scope of the present application to any single invention or inventive concept if more than one invention or inventive concept is in fact disclosed. Therefore, although specific embodiments have been shown and described herein, it should be understood that any arrangement intended to achieve the same purpose may replace the specific embodiments shown. The present disclosure is intended to cover any and all modifications or variations of the various embodiments. After reading the above description, combinations of the above embodiments and other embodiments not specifically described herein will be obvious to those skilled in the art.

[0090] Example

[0091] 1. A computer-implemented method comprising:

[0092] receiving, by a base recommendation engine implemented using at least one hardware processor, a recommendation request from a computing device;

[0093] generating, by the base recommendation engine, a first plurality of recommendations based on the recommendation request using a base recommendation model;

[0094] Obtaining, by the base recommendation engine, one or more corresponding sets of other recommendations from one or more other recommendation engines, each of the one or more other recommendation engines being configured to generate one or more corresponding sets of other recommendations based on the recommendation request using a corresponding recommendation model different from the base recommendation model;

[0095] For each of the one or more other recommendation engines, the basic recommendation engine obtains the corresponding weight of the other recommendation engine from a database;

[0096] selecting, by the base recommendation engine, at least one recommendation from the first plurality of recommendations and each corresponding set of one or more other recommendations from the one or more other recommendation engines using each corresponding weight of each set of one or more other recommendations from the one or more other recommendation engines; and

[0097] The selected at least one recommendation is caused to be displayed on a computing device by the base recommendation engine.

[0098] 2. The computer-implemented method of example 1, further comprising:

[0099] receiving corresponding feedback data for each of the selected at least one recommendation, the feedback data indicating a reaction of a user of the computing device to display of the selected at least one recommendation on the computing device; and

[0100] The received feedback data is used as training data in at least one machine learning operation to modify a corresponding weight of at least one other recommendation engine of the one or more other recommendation engines stored in a database.

[0101] 3. A computer-implemented method according to Example 1 or Example 2, wherein the base recommendation engine and each of the one or more other recommendation engines form a group of recommendation engines, each recommendation engine being configured to serve a specific category of organizational functions that is different from the specific category of organizational functions of all other recommendation engines in the group.

[0102] 4. A computer-implemented method according to any one of Examples 1 to 3, wherein the base recommendation engine and each of the one or more other recommendation engines form a group of recommendation engines, each recommendation engine having its own application domain that is different from the application domains of all other recommendation engines in the group.

[0103] 5. A computer-implemented method according to any one of Examples 1 to 4, wherein, for each of the one or more other recommendation engines, the corresponding set of one or more other recommendations is obtained from the one or more other recommendation engines via a direct network connection between the base recommendation engine and the one or more other recommendation engines.

[0104] 6. The computer-implemented method of any one of Examples 1 to 5, wherein, for each of the one or more other recommendation engines, the corresponding set of one or more other recommendations is obtained from the one or more other recommendation engines via a management server separate and distinct from the base recommendation engine, the management server being configured to:

[0105] Retrieve said corresponding set of one or more other recommendations from each of said one or more other recommendation engines; and

[0106] The corresponding set of one or more other recommendations obtained from each of the one or more other recommendation engines is sent to the base recommendation engine.

[0107] 7. The computer-implemented method of any one of Examples 1 to 6, wherein the base recommendation engine is configured to:

[0108] for each of the one or more other recommendation engines, obtaining, via a direct network connection between the base recommendation engine and the one or more other recommendation engines, the corresponding set of one or more other recommendations from the one or more other recommendation engines;

[0109] detecting a switching event corresponding to one of the one or more other recommendation engines; and

[0110] Based on the detected switching event, obtaining a corresponding set of one or more other recommendations from other recommendation engines corresponding to the switching event via a management server separate and different from the basic recommendation engine, the management server being configured to:

[0111] Obtaining a corresponding set of one or more other recommendations from other recommendation engines corresponding to the switching event; and

[0112] A corresponding set of one or more other recommendations obtained from other recommendation engines corresponding to the switching event is sent to the basic recommendation engine.

[0113] 8. The computer-implemented method according to any one of Examples 1 to 7 further includes selecting one or more other recommendation engines from a plurality of recommendation engines to be used to obtain the corresponding set of one or more other recommendations based on the corresponding weight of each of the selected one or more other recommendation engines.

[0114] 9. A computer-implemented method according to Example 8, wherein selecting the one or more other recommendation engines includes omitting at least one of the multiple recommendation engines from the selection based on a corresponding weight of the at least one of the multiple recommendation engines.

[0115] 10. The computer-implemented method of any one of Examples 1 to 9, wherein the selected at least one recommendation comprises at least one recommendation for online content or at least one recommendation for performing an action.

[0116] 11. A system comprising:

[0117] at least one processor; and

[0118] A non-transitory computer-readable medium storing executable instructions that, when executed, cause the at least one processor to perform the method according to any one of Examples 1 to 10.

[0119] 12. A non-transitory machine-readable storage medium tangibly embodying a set of instructions that, when executed by at least one processor, cause the at least one processor to perform the method according to any one of Examples 1 to 10.

[0120] 13. A machine-readable medium carrying an instruction set, which, when executed by at least one processor, causes the at least one processor to perform a method according to any one of Examples 1 to 10.

[0121] The abstract of the present disclosure is provided to allow the reader to quickly determine the nature of the technical disclosure. The abstract is submitted with the understanding that it will not be used to interpret or limit the scope or meaning of the claims. In addition, in the foregoing detailed description, it can be seen that various features are grouped together in a single embodiment in order to simplify the present disclosure. The method of the present disclosure should not be interpreted as reflecting the intention that the claimed embodiments require more features than those expressly stated in each claim. On the contrary, as reflected in the attached claims, the inventive subject matter lies in less than all the features of a single disclosed embodiment. Therefore, the attached claims are incorporated into the detailed description, with each claim independently serving as a separate embodiment.

Claims

1. A computer-implemented method comprising: receiving, by a base recommendation engine implemented using at least one hardware processor, a recommendation request from a computing device; generating, by the base recommendation engine, a first plurality of recommendations based on the recommendation request using a base recommendation model; The basic recommendation engine obtains a corresponding set of one or more other recommendations from one or more other recommendation engines, each of the one or more other recommendation engines being configured to generate the corresponding set of one or more other recommendations based on the recommendation request using a corresponding recommendation model different from the basic recommendation model, and the basic recommendation engine being configured to: for each of the one or more other recommendation engines, obtaining the corresponding set of one or more other recommendations from the one or more other recommendation engines via a direct network connection between the base recommendation engine and the one or more other recommendation engines; detecting a switching event corresponding to one of the one or more other recommendation engines, the switching event comprising a network disconnection between the base recommendation engine and the one of the one or more other recommendation engines; and Based on the detected switching event, obtaining the corresponding set of one or more other recommendations from other recommendation engines corresponding to the switching event via a management server separate and different from the basic recommendation engine; selecting, by the base recommendation engine, at least one recommendation from each corresponding set of the first plurality of recommendations and one or more other recommendations from the one or more other recommendation engines; and The selected at least one recommendation is caused to be displayed on the computing device by the base recommendation engine.

2. The computer-implemented method of claim 1 , further comprising: for each of the one or more other recommendation engines, obtaining, by the base recommendation engine, corresponding weights of the other recommendation engines from a database, wherein selecting the at least one recommendation is performed using each corresponding weight of each set of one or more other recommendations from the one or more other recommendation engines; receiving corresponding feedback data for each of the selected at least one recommendation, the feedback data indicating a reaction of a user of the computing device to display of the selected at least one recommendation on the computing device; and The received feedback data is used as training data in at least one machine learning operation to modify a corresponding weight of at least one other recommendation engine of the one or more other recommendation engines stored in a database.

3. The computer-implemented method of claim 1 , wherein: The base recommendation engine and each of the one or more other recommendation engines form a group of recommendation engines, and each recommendation engine in the group of recommendation engines is configured to serve a specific category of organizational functions, which is different from the specific category of organizational functions of all other recommendation engines in the group.

4. The computer-implemented method of claim 1 , wherein: The base recommendation engine and each of the one or more other recommendation engines form a group of recommendation engines, each recommendation engine in the group of recommendation engines having their own application domain that is different from application domains of all other recommendation engines in the group.

5. The computer-implemented method of claim 1 , wherein: For each of the one or more other recommendation engines, the corresponding set of one or more other recommendations is obtained from the one or more other recommendation engines via a direct network connection between the base recommendation engine and the one or more other recommendation engines.

6. The computer-implemented method of claim 1, wherein: For each of the one or more other recommendation engines, obtaining the corresponding set of one or more other recommendations from the one or more other recommendation engines via a management server separate and distinct from the base recommendation engine, the management server being configured to: obtaining the corresponding set of one or more other recommendations from each of the one or more other recommendation engines; and The corresponding set of one or more other recommendations obtained from each of the one or more other recommendation engines is sent to the base recommendation engine.

7. The computer-implemented method of claim 1 , wherein: The management server is configured as: Obtain the corresponding set of one or more other recommendations from other recommendation engines corresponding to the switching event; and The corresponding set of one or more other recommendations obtained from other recommendation engines corresponding to the switching event is sent to the basic recommendation engine.

8. The computer-implemented method of claim 1 , further comprising: for each of the one or more other recommendation engines, obtaining, by the base recommendation engine, corresponding weights of the other recommendation engines from a database, wherein selecting the at least one recommendation is performed using each corresponding weight of each set of one or more other recommendations from the one or more other recommendation engines; as well as Based on the corresponding weight of each of the selected one or more other recommendation engines, the one or more other recommendation engines are selected from a plurality of recommendation engines to be used to obtain the corresponding set of one or more other recommendations.

9. The computer-implemented method of claim 8, further comprising: For each of the one or more other recommendation engines, the basic recommendation engine obtains the corresponding weights of the other recommendation engines from the database, wherein selecting the at least one recommendation is performed using each corresponding weight of each set of one or more other recommendations from the one or more other recommendation engines, wherein selecting the one or more other recommendation engines includes omitting at least one of the multiple recommendation engines from the selection based on the corresponding weight of at least one of the multiple recommendation engines.

10. The computer-implemented method of claim 1, wherein: The selected at least one recommendation includes at least one recommendation for online content or at least one recommendation for performing an action.

11. A system for implementing artificial swarm intelligence via networking of recommendation engines, comprising: at least one hardware processor; and A non-transitory computer-readable medium storing executable instructions that, when executed, cause the at least one processor to perform operations comprising: receiving, by a base recommendation engine implemented using the at least one hardware processor, a recommendation request from a computing device; generating, by the base recommendation engine, a first plurality of recommendations based on the recommendation request using a base recommendation model; The basic recommendation engine obtains a corresponding set of one or more other recommendations from one or more other recommendation engines, each of the one or more other recommendation engines being configured to generate the corresponding set of one or more other recommendations based on the recommendation request using a corresponding recommendation model different from the basic recommendation model, and the basic recommendation engine being configured to: for each of the one or more other recommendation engines, obtaining the corresponding set of one or more other recommendations from the one or more other recommendation engines via a direct network connection between the base recommendation engine and the one or more other recommendation engines; detecting a switching event corresponding to one of the one or more other recommendation engines, the switching event comprising a network disconnection between the base recommendation engine and the one of the one or more other recommendation engines; and Based on the detected switching event, obtaining the corresponding set of one or more other recommendations from other recommendation engines corresponding to the switching event via a management server separate and different from the basic recommendation engine; selecting, by the base recommendation engine, at least one recommendation from each corresponding set of the first plurality of recommendations and one or more other recommendations from the one or more other recommendation engines; and The selected at least one recommendation is caused to be displayed on the computing device by the base recommendation engine.

12. The system according to claim 11, wherein: The operations also include: for each of the one or more other recommendation engines, obtaining, by the base recommendation engine, corresponding weights of the other recommendation engines from a database, wherein selecting the at least one recommendation is performed using each corresponding weight of each set of one or more other recommendations from the one or more other recommendation engines, receiving corresponding feedback data for each of the selected at least one recommendation, the feedback data indicating a reaction of a user of the computing device to display of the selected at least one recommendation on the computing device; and The received feedback data is used as training data in at least one machine learning operation to modify a corresponding weight of at least one other recommendation engine of the one or more other recommendation engines stored in a database.

13. The system according to claim 11, wherein: The base recommendation engine and each of the one or more other recommendation engines form a group of recommendation engines, and each recommendation engine in the group of recommendation engines is configured to serve a specific category of organizational functions, which is different from the specific category of organizational functions of all other recommendation engines in the group.

14. The system according to claim 11, wherein: The base recommendation engine and each of the one or more other recommendation engines form a group of recommendation engines, each recommendation engine in the group of recommendation engines having their own application domain that is different from application domains of all other recommendation engines in the group.

15. The system according to claim 11, wherein: For each of the one or more other recommendation engines, the corresponding set of one or more other recommendations is obtained from the one or more other recommendation engines via a direct network connection between the base recommendation engine and the one or more other recommendation engines.

16. The system according to claim 11, wherein: For each of the one or more other recommendation engines, obtaining the corresponding set of one or more other recommendations from the one or more other recommendation engines via a management server separate and distinct from the base recommendation engine, the management server being configured to: obtaining the corresponding set of one or more other recommendations from each of the one or more other recommendation engines; and The corresponding set of one or more other recommendations obtained from each of the one or more other recommendation engines is sent to the base recommendation engine.

17. The system of claim 11, wherein: The management server is configured as: Obtain the corresponding set of one or more other recommendations from other recommendation engines corresponding to the switching event; and The corresponding set of one or more other recommendations obtained from other recommendation engines corresponding to the switching event is sent to the basic recommendation engine.

18. The system of claim 11, wherein: The operations also include: for each of the one or more other recommendation engines, obtaining, by the base recommendation engine, corresponding weights of the other recommendation engines from a database, wherein selecting the at least one recommendation is performed using each corresponding weight of each set of one or more other recommendations from the one or more other recommendation engines, and Based on the corresponding weight of each of the selected one or more other recommendation engines, the one or more other recommendation engines are selected from a plurality of recommendation engines to be used to obtain the corresponding set of one or more other recommendations.

19. The system of claim 18, wherein: The operations also include: For each of the one or more other recommendation engines, the base recommendation engine obtains the corresponding weights of the other recommendation engines from the database, wherein selecting the at least one recommendation is performed using each corresponding weight of each set of one or more other recommendations from the one or more other recommendation engines, and wherein selecting the one or more other recommendation engines includes omitting at least one of the multiple recommendation engines from the selection based on the corresponding weight of at least one of the multiple recommendation engines.

20. A non-transitory machine-readable storage medium tangibly embodying a set of instructions that, when executed by at least one processor, cause the at least one processor to perform operations comprising: receiving, by a base recommendation engine implemented using the at least one hardware processor, a recommendation request from a computing device; generating, by the base recommendation engine, a first plurality of recommendations based on the recommendation request using a base recommendation model; The basic recommendation engine obtains a corresponding set of one or more other recommendations from one or more other recommendation engines, each of the one or more other recommendation engines being configured to generate the corresponding set of one or more other recommendations based on the recommendation request using a corresponding recommendation model different from the basic recommendation model, and the basic recommendation engine being configured to: for each of the one or more other recommendation engines, obtaining the corresponding set of one or more other recommendations from the one or more other recommendation engines via a direct network connection between the base recommendation engine and the one or more other recommendation engines; detecting a switching event corresponding to one of the one or more other recommendation engines, the switching event comprising a network disconnection between the base recommendation engine and the one of the one or more other recommendation engines; and Based on the detected switching event, obtaining the corresponding set of one or more other recommendations from other recommendation engines corresponding to the switching event via a management server separate and different from the basic recommendation engine; selecting, by the base recommendation engine, at least one recommendation from each corresponding set of the first plurality of recommendations and one or more other recommendations from the one or more other recommendation engines; and The selected at least one recommendation is caused to be displayed on the computing device by the base recommendation engine.