Generative artificial intelligence knowledge graph engine in item listing system

Abstract

Methods, systems, and computer storage media for providing knowledge graphs using a Generative AI Knowledge Graph (KG) engine "Generative AI KG Engine" in an item listing system. A generative AI KG engine supports the generation of knowledge graphs using a generative AI model (e.g., LLM). In operation, a seed product is accessed in a product listing system. Using the product knowledge graph, a plurality of candidate products associated with the seed product are identified. The product knowledge graph includes a plurality of products as nodes and a plurality of relationships as edges, the product knowledge graph being associated with a generative AI model. A plurality of recommended products are identified using a ranker of the product listing system. The plurality of recommended products is a subset of the plurality of candidate products. A plurality of recommended products are transmitted, and the plurality of recommended items are caused to be generated on a graphical user interface.

Description

Technical Field

[0001] This application relates to a generative artificial intelligence knowledge graph engine in a project list system. Background Art

[0002] Users can interact with generative artificial intelligence technology in different types of applications and services to complete computational tasks. Generative AI refers to a category of AI systems and algorithms that are designed to generate new data or content that is similar to or, in some cases, completely different from the data on which they are trained. Generative AI systems can create support for text generation, image generation, music and audio generation, video generation, and data synthesis. Specifically, generative AI systems can support a project list system in several ways to improve operational efficiency, customer engagement, and online shopping. For example, a project list system can employ a generative AI system for content generation (e.g., product descriptions), personalized shopping experiences (e.g., recommendation engines), product discovery (e.g., visual search), and virtual assistants (e.g., chatbots). A project list system can utilize generative AI to enhance project list functionality through application programming interfaces (APIs), pre-trained models, and custom AI solutions. Summary of the Invention

[0003] Aspects of the technology described herein generally relate to systems, methods, and computer storage media for using an artificial intelligence system associated with a project list system to provide a knowledge graph and the like. The generative AI knowledge graph engine of the artificial intelligence system, the "Generative AI KG Engine", supports providing a knowledge graph in the project list system. The Generative AI KG Engine supports using a generative AI model (e.g., a large language model) to generate a knowledge graph (e.g., a product knowledge graph or a Generative AI KG). Specifically, the generative AI knowledge graph (KG) engine includes an edge-node prediction service - wherein nodes and edges of the knowledge graph are generated based on data (e.g., one or more prompts executed on the generative AI model) generated from a product knowledge graph. The Generative AI KG Engine also includes a mapping service that maps products (with KG information) to products in the product list database (e.g., a project list database) of the project list system. A product knowledge graph or a mapped product knowledge graph can be deployed to support different types of services (e.g., recommendations, product search, seller feedback services) in the project list system.

[0004] In operation, a product listing system with a seed product is provided. The seed product is accessed. Using a product knowledge graph, a plurality of candidate products associated with the seed product are identified. The product knowledge graph includes a plurality of products as nodes and a plurality of relationships as edges, and the product knowledge graph is associated with a generative AI model. A plurality of recommended products are identified. A ranker of the product listing system can be used to identify the plurality of recommended products. The plurality of recommended products is a subset of the plurality of candidate products. The plurality of recommended products is transmitted and caused to be generated on a graphical user interface.

[0005] The present invention content is provided to introduce, in a simplified form, a selection of concepts that are further described below in the detailed description. The present invention content is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to assist in determining the scope of the claimed subject matter. BRIEF DESCRIPTION OF THE DRAWINGS

[0006] The techniques described herein are described in detail below with reference to the accompanying drawings, in which:

[0007] Figure 1A and Figure 1B is a block diagram of an artificial intelligence system for providing a product knowledge graph function in a project listing system according to aspects of the techniques described herein;

[0008] Figures 1C to 1F is a schematic diagram and interface associated with providing a product knowledge graph function in a project listing system according to aspects of the techniques described herein;

[0009] Figure 2A is a block diagram of an artificial intelligence system for providing a product knowledge graph function in a project listing system according to aspects of the techniques described herein;

[0010] Figure 2B is a block diagram of an artificial intelligence system for providing a product knowledge graph function in a project listing system according to aspects of the techniques described herein;

[0011] Figure 3 A first exemplary method for providing a product knowledge graph function in a project listing system according to aspects of the techniques described herein is provided;

[0012] Figure 4 A second exemplary method for providing a product knowledge graph function in a project listing system according to aspects of the techniques described herein is provided;

[0013] Figure 5 A third exemplary method for providing a product knowledge graph function in a project listing system according to aspects of the techniques described herein is provided;

[0014] Figure 6Provides a block diagram of an exemplary project listing system computing environment suitable for implementing the technical aspects described herein;

[0015] Figure 7 Provides a block diagram of an exemplary distributed computing environment suitable for implementing the technical aspects described herein; and

[0016] Figure 8 Is a block diagram of an exemplary computing environment suitable for implementing the technical aspects described herein. Detailed Description

[0017] Overview

[0018] A project listing system and platform support storing projects (products or assets) in a project database and provide a search system for receiving queries and identifying search result projects based on the query. A project (e.g., a physical project or a digital project) refers to a product or asset offered for listing on a project listing platform. The search system supports identifying result projects from the project database for the received query. The project database can be dedicated to a content platform or a project listing platform, such as the EBAY content platform developed by EBAY Inc. in San Jose, California. The project listing system can also provide generative AI-supported applications (“generative AI applications”) that utilize generative AI models (e.g., image generation models and large language models - “LLMs”) to create, generate, or produce content, data, or output. An LLM is a specific category of generative AI model that mainly focuses on generating human-like text. Generative AI models (such as GPT (Generative Pretrained Transformer) and its variants) are designed to generate human-like text or other types of data based on the input they receive (e.g., via a prompt interface). These applications use generative AI to perform various tasks across different domains to provide improvements in automation, efficiency, and human-like interaction.

[0019] Traditionally, project listing systems have not been configured with comprehensive logic and infrastructure to effectively provide a generative artificial intelligence-based product knowledge graph for the project listing system. Conventional project listing systems have limited understanding of products and do not include data structures with a comprehensive understanding of products. In addition, existing data structures are not easily integrated into different types of generative AI-supported functions (e.g., search, recommendation, product discovery) of the project listing system. Project listing systems can rely on manual tagging of products, where product relationships are annotated by human experts. Manual methods of understanding product relationships can be very time-consuming and labor-intensive. In addition, it may also limit the ability to identify and utilize a comprehensive set of product relationships for promotional strategies or other functions (e.g., search or seller feedback) in the project listing system.

[0020] Implementing only a conventional knowledge graph (without a generative AI knowledge graph (KG) engine) would result in insufficient functionality of the item listing system. For example, traditional knowledge graphs may face a series of challenges, ranging from data integrity and accuracy to limited context information. Specifically, knowledge graphs rely on accurate and complete data; if the underlying product data is incomplete, outdated, or inaccurate, it may lead to an incomplete or incorrect representation of the knowledge graph. The knowledge graph may lack the ability to capture fine-grained context information about products (e.g., user preferences, real-time popularity, or temporal changes in product features). The knowledge graph may encounter the "cold start" problem, where new products or categories lack sufficient historical data. This makes it challenging to provide accurate recommendations for recently launched items.

[0021] The conventional item listing system can be improved by addressing these limitations through the use of advanced machine learning models and techniques - these advanced machine learning models and techniques can enhance the flexibility and adaptability of the product knowledge graph in capturing complex relationships and changing user preferences. Thus, a more comprehensive item listing system (with an alternative foundation for performing the operations of a generative AI knowledge graph (KG) engine) can improve the computational operations and interfaces for providing item listing services in the item listing system.

[0022] Embodiments of the present invention relate to systems, methods, and computer storage media for using an artificial intelligence system associated with an item listing system to provide a knowledge graph, etc. The generative AI knowledge graph engine of the artificial intelligence system, the "generative AI KG engine", supports providing a knowledge graph in the item listing system. The generative AI KG engine supports using a generative AI model (e.g., a large language model) to generate a knowledge graph (e.g., a product knowledge graph or a generative AI KG). Specifically, the generative AI knowledge graph (KG) engine includes an edge-node prediction service - where data (e.g., one or more prompts executed on the generative AI model) is generated based on the product knowledge graph to generate the nodes and edges of the knowledge graph.

[0023] The generative AI KG engine also includes a mapping service that maps products (with KG information) to products in the product listing database (e.g., the item listing database) of the item listing service. The product knowledge graph or the mapped product knowledge graph can be deployed to support different types of services in the item listing system (e.g., recommendations, product search, seller feedback services).

[0024] The generative AI KG is provided using a generative AI KG engine that is operationally integrated into the item listing system associated with the artificial intelligence system. The artificial intelligence system supports the generative AI KG framework of the computational components associated with the operation of the generative AI KG engine for providing the generative AI KG.

[0025] At a high level, understanding the relationships between products is very important in a project listing system or a product listing system. Specifically, offering affordable alternative products or complementary products to customers can be based on the understanding of the relationships between products. For example, a customer who may be hesitating to purchase a high-priced product may receive a recommendation for a mid-priced alternative product and decide to purchase the alternative product. Additionally, another customer may receive a recommendation for a complementary product after purchasing a related product. However, capturing product relationships can be a challenging task.

[0026] For example, existing recommendation systems in a project listing system may focus on behavioral data; however, with generative AI models (e.g., LLM), a broader volume of data can be accessed to improve the recommendation system and other types of functionality in the project listing system. The LLM can store knowledge from different domains and fields that further complements the limited behavioral data that the recommendation system previously relied on. For example, an LLM-based system such as ChatGPT may recommend purchasing a turkey on Thanksgiving in a zero-shot manner (i.e., without prior information or historical data and without user preferences or interactions), even without click behavioral data related to turkeys or Thanksgiving. Nevertheless, the project listing system has not fully integrated the LLM into the functions (e.g., applications and services) of the project listing system in the most effective and efficient way. Specifically, latency and cost have been considerations that limit the integration of the LLM into the project listing system.

[0027] The generative AI knowledge graph framework is a technical solution that involves using a generative AI engine of a project listing system to provide a knowledge graph. The generative AI knowledge graph framework includes a generative AI KG engine that supports using a generative AI model (e.g., large language model - "LLM") to generate a knowledge graph (e.g., product knowledge graph or generative AI KG). For example, the LLM can provide natural language processing (NLP) based on pre-trained models that encode a large amount of information. The LLM can be utilized to understand the content or items (e.g., products) of the project listing platform to generate an improved representation of the relationships between items. Specifically, the LLM can be applied to multiple downstream NLP applications.

[0028] An LLM can be associated with seed input prompts (e.g., product knowledge graph generation data) - these seed input prompts are knowledge graph generation prompts that support determining relationships between products (e.g., a first product and a second product; or a primary product and a complementary product). Based on executing the seed input prompts on the LLM, the LLM results can be used to generate a product knowledge graph. The product knowledge graph improves latency issues and costs associated with other integration methods of the LLM in a project list system. The product knowledge graph can be deployed and integrated (e.g., via an application programming interface) into different types of software (e.g., applications and services - collectively referred to as "applications") in the project list system.

[0029] In operation, a project list system (e.g., a product list system) can implement an artificial intelligence system that provides a generative artificial intelligence knowledge graph engine. The generative AI KG engine operates as an edge-node prediction service that uses a generative AI model to generate a knowledge graph. The generative AI KG engine operates with a project list system client (e.g., a product list system client) to provide the functions of the project list system based on the knowledge graph (e.g., a generative AI knowledge graph or a product knowledge graph).

[0030] The knowledge graph is generated using a machine learning engine, a generative AI model, and product knowledge graph generation data. The product knowledge graph generation data can include data associated with products in the project list system, where the product knowledge graph generation data includes seed prompt inputs based on a prompt template that support generating the product knowledge graph. The generative AI model can be an LLM (e.g., a generative pre-trained transformer model) for generating the product knowledge graph. The generative AI model can be specifically trained using knowledge graph generation training operations and product knowledge graph generation data that support generating the product knowledge graph.

[0031] The product knowledge graph can be transformed into a mapped product knowledge graph using mapping rules and mapping operations (e.g., knowledge graph mapping operations) that support mapping products to products in the project list system. The mapping rules can support mapping products to the same products or similar products in the project list system. The project list system (e.g., the project database in the project list system or the product list system database) can include multiple instances of the same product, so a product can be specifically mapped to a particular instance of a product in the project list system based on the characteristics of the node associated with the product in the product knowledge graph and the characteristics of the product instance on the product list system.

[0032] A product knowledge graph (or a mapped product knowledge graph) can be integrated into a project listing system service that uses the product knowledge graph to perform product listing system functions. For example, the project listing system service can be a search engine or a recommendation service that uses the product knowledge graph to provide search result products or recommended products. For example, the project listing system service can use a ranker of the project listing system to rank one or more products identified using the product knowledge graph and the project listing system database to identify a subset of products to be transmitted as search result products or recommended products. The product knowledge graph can be used to identify products displayed on different types of interfaces associated with the project listing system (e.g., a seller interface, a buyer interface, or a recommendation interface).

[0033] Advantageously, embodiments of the present technical solution support using a generative AI KG engine in a project listing system to provide a knowledge graph. The generative AI KG engine supports using a generative AI model (e.g., a large language model) to generate a knowledge graph (e.g., a product knowledge graph or a generative AI KG). The generative AI KG engine operates to provide a solution to problems in the project listing system (e.g., limited ability to identify and utilize a comprehensive set of product relationships for project listing system services and functions). The combination of generative AI KG engine components, infrastructure, and ordered steps is an improvement over conventional project listing systems that lack support for generative AI knowledge graphs.

[0034] As an example and with reference to Figures 1A to 1F the aspects of the technical solution can be described. Figure 1A Project listing system 100 is shown, which includes an artificial intelligence system 100A, a network 100B, a generative artificial intelligence (AI) knowledge graph engine 110, a project listing system service 110B, a knowledge graph integration API 110C, a generative AI application 110D, a project listing system client 130, and a machine learning engine 140 including a generative AI model 142. Project listing system 100 (or product listing system) corresponds to project listing system 600 described below with reference to Figure 6 the above.

[0035] The item listing system 100 provides a system (e.g., an artificial intelligence "AI" system 100A) that includes an engine (e.g., a generative AI KG engine 110) for performing the operations described herein (e.g., knowledge graph engine operations). The generative AI KG engine 110 can operate in conjunction with an item listing system client 130 (e.g., a client device or a generative AI application client), which can access the item listing system 100 to perform tasks using a generative AI application 110D associated with a corresponding generative AI model 142. For example, via the item listing system client 130 (e.g., a prompt interface), a user can transmit a request (e.g., a generative AI request with prompt data) to the generative AI application 110D and the generative AI model 142 associated with the machine learning engine 140 to process the request. Based on transmitting the request, the generative AI KG engine 110 can perform knowledge graph engine operations (e.g., training, generation, deployment, integration, mapping, prediction, and control operations) using components of the generative AI KG engine 110 to ensure processing of the request.

[0036] The generative AI KG engine 110 can also include item listing system services 110B corresponding to different services of the item listing system 100. The item listing system services 110B can include a search service and a recommendation service, which, for example, use a product knowledge graph 120 to provide item listing system functionality. The item listing system services 110B can include a product listing system service and a ranker associated with the item listing system 100. A knowledge graph integration application programming interface (API) 110C can be provided to integrate the item listing system services 110B with the product knowledge graph 120. The generative AI application 110D can also operate to use the product knowledge graph 120 to provide functionality associated with the generative AI application. The embodiments described herein contemplate other variations and combinations of item listing system services.

[0037] Reference Figure 1B , Figure 1B illustrates the item listing system 100, the artificial intelligence system 100A, the generative AI KG engine 110, the product knowledge graph generation data 110D, the knowledge graph engine operations 112, the knowledge graph data structure 114, the edge-node prediction service 116, the knowledge graph mapping operations and rules 118, the product knowledge graph 120, the mapped product knowledge graph 122, the item listing system client 130 including item listing system client interface data 132, the machine learning engine 140 including the generative AI model 142, and the item listing system services 110B including the product listing system service 112B, the ranker 114B, and the knowledge graph integration API 110C.

[0038] The Generative AI KG Engine 110 and the Project List System Client 130 provide a graphical user interface (e.g., the Project List System Interface and the Generative AI Application Interface) and operations (i.e., Knowledge Graph Engine operations). The Generative AI KG Engine 110 and the Project List System Client 130 can operate in a server-client relationship to provide Product Knowledge Graph (e.g., Product Knowledge Graph 120 and Mapped Product Knowledge Graph 122) functionality. For example, a user can transmit a request from the Project List System Client 130 to perform a task via the Generative AI Application and the Product Knowledge Graph 120 or the Mapped Product Knowledge Graph 122. Based on the request, the Generative AI KG Engine 110 can perform Knowledge Graph Engine Operation 112 to ensure that the request is processed in the Artificial Intelligence System 100A.

[0039] The Generative AI KG Engine 110 can perform Knowledge Graph Engine Operation 110A to provide functionality associated with generating, deploying, integrating, and using the Product Knowledge Graph 120. Knowledge Graph Engine Operation 112 can support prompt engineering. Specifically, the Generative AI KG Engine 110 can include instructions for the prompt design of the Generative AI Model 142 such that the prompt is constructed to provide a response from the Generative AI Model 142 to the Product Knowledge Graph 120. The Generative AI KG Engine 110 can design and provide precise language, context cues, task specifications, and examples and demonstrations in the prompt template. Additional functionality can be associated with fine-tuning and iteration to handle ambiguity, model bias, and balance the openness and control of the Generative AI Model 142.

[0040] The Product Knowledge Graph Generation Data 110D can include a prompt template for generating a prompt to be executed on the Generative AI Model 142. The Product Knowledge Graph Generation Data 110D can be processed using prompt engineering operations associated with the Knowledge Graph Engine Operation 112. Specifically, prompt engineering can support generating a template, which can be converted into a prompt based on different types of data from the Product Knowledge Graph Data 110D. The template can include a template format that can be filled with information from the Product Knowledge Graph Data 110D. For example, the template can include context filled based on the title of the seed product, historical user behavior, and interface user responses. The prompt generated using the Prompt Knowledge Graph Generation Data 110D can be executed on the Generative AI Model in batch mode to simplify the execution of multiple prompts associated with generating the Product Knowledge Graph. The prompt template can include one or more features disclosed in Table 1 below. Table 1 - Prompt Template Features

[0041] Two example templates are shown below.

[0042] A product knowledge graph 120 can be generated based on executing a prompt on a generative AI model 142. The product knowledge graph 120 can be a structured representation of knowledge that captures relationships between entities (i.e., products in a product listing system). It consists of nodes and edges, where nodes represent entities and edges represent relationships between these entities. Nodes are entities in the knowledge graph. Each node has attributes or characteristics associated with it, providing additional information about the entity. Example product nodes can be associated with the following: product title - Nike Air Force 1; product type - sports shoes; product brand - Nike; product audience - sports shoe enthusiasts, sports lovers, fashionistas, and young people (aged 15 to 35).

[0043] Edges represent relationships between nodes in the knowledge graph. These relationships define how different entities are connected or associated with each other. Edges have labels that describe the nature of the relationship. Example edges can be associated with the edge topic "Nike Air Force 1"; edge predicate - choice of fashionistas; and edge object: Nike Heritage Backpack. The output of the generative AI model can be processed to map the product knowledge graph to an item listing database (e.g., the inventory of an item listing system). Mapping operations and mapping rules can be used to process the output.

[0044] Knowledge graph mapping operations and rules 118 can be implemented to transform the product knowledge graph into a mapped product knowledge graph 122. Specifically, products in the product knowledge graph 120 are mapped to corresponding products in the product listing database (e.g., exact matches or partial matches). Exact matches and partial matches can be based on mapping rules. Mapping rules can include matching rules for identifying and exactly mapping as well as partially mapping closely related products between the product knowledge graph and the product listing system database. These rules include category matching, where products within the same or similar categories are considered related; brand matching, associating products from the same brand; attribute matching, focusing on products with similar specifications or characteristics; price range matching, recommending products within a comparable price range; usage context matching, suggesting products that are commonly used together.

[0045] In addition, historical purchase matching leverages user behavior, while customer preference matching aligns recommendations with individual user preferences. Seasonal matching provides relevant suggestions based on the current season or upcoming events, while accessory matching suggests complementary items. Popularity matching considers trending or frequently purchased products, and geographic matching customizes recommendations based on regional relevance.

[0046] In one embodiment, a machine learning model can be trained to predict product relationships based on historical data. For example, a collaborative filtering model can learn from past user behavior to predict which products are likely to be relevant. As mentioned previously (category matching, brand matching, etc.), the matching rules are implemented in the form of logical conditions within the algorithm. These rules guide the system to identify closely related products. The algorithm assigns a score or ranking to them based on the degree to which the potential matches satisfy the matching rules. For example, products in the same category can receive a higher score. The embodiments described herein consider other variants and combinations of the rules, which are mapping rules, matching rules, and implementation methods.

[0047] The item listing system service 110B can include different types of services provided by the item listing system (e.g., product recommendation applications, product search applications, and seller feedback services). For example, the product listing system service 112 can be associated with sellers to support listing products for sale. The ranker 114B supports a ranking function as a complex system within the item listing system to orchestrate the display of products to users based on a multi-faceted scoring mechanism. For example, by using the relevance score of each product, the ranker combines factors such as user engagement metrics, historical behavior patterns, and product attributes to dynamically identify the items most suitable for an individual user.

[0048] By leveraging advanced machine learning models (including but not limited to collaborative filtering and content-based filtering), the ranker can predict user preferences and generate personalized recommendations. User behavior analysis combines a series of interactions, including clicks, purchases, and temporal dynamics, to help the system adapt over time. In addition, the ranker integrates semantic understanding and context information, including location, device type, and real-time user behavior, to highly customize recommendations. The item listing system service can be integrated with a product knowledge graph to support the corresponding functions of the item listing system service.

[0049] The item listing system client 130 can be associated with a seller interface, a buyer interface, and other item listing system service interfaces associated with the item listing system. The item listing system client 130 can cause the display of item listing system client interface data 132 associated with products (e.g., search result products or recommended products) associated with the item listing system 100 based on the generative AI KG engine 110, the generative AI model 142, the product knowledge graph 120, and functions associated with the item listing system 100. The item listing system client interface data can be associated with different outputs corresponding to the item listing system service 110B and other functional components, as well as the output of the generative AI KG engine 110D.

[0050] Turning to Figure 1C , Figure 1CShows a schematic diagram associated with providing a knowledge graph using a generative AI KG engine according to embodiments described herein. Figure 1C Includes a generative AI knowledge graph framework, which includes product 112C, LLM 114C with edge-node prediction service 116C, product knowledge graph 120C, product 122C, edge 124C_1, edge 124C_2, node attribute 126C, edge attribute 128C, mapping rule 130C (including nodes mapped to product 132C), product-to-product API 140C, and graph database 150C.

[0051] For example, product 112C (e.g., a seed product) can be processed via a generative AI model (e.g., LLM 114C) that provides the functionality of edge-node prediction service 116C. Product 112C can be provided as product knowledge graph generation data in a prompt template designed to elicit a response from the LLM. Edge-node prediction service 116 makes predictions or inferences on the products and the relationships between products in product knowledge graph 120 on the LLM.

[0052] Product knowledge graph 120 includes multiple products (e.g., product 122C_1 and product 122C_2). Products are connected to other products via edges (e.g., edge 124C_1 and edge 124_C2). Products can be unique products identified using LLM 114C, and the edges are based on relationships defined by the prompt (e.g., similar style, affordable alternatives, and compatibility). Products are nodes associated with node attributes (e.g., product ID, category, product title, and audience), and relationships are edges associated with edge attributes (e.g., relationship and relationship type).

[0053] Multiple mapping rules are provided to map products to products in a product list database. Specifically, the product list database can be a set of products with product identifiers (e.g., Nike Air Force 1’07). Each instance of a product is associated with a product identifier, but each instance can be associated with additional features (e.g., seller, cost, color, etc.). If the product in the product knowledge graph is the same as the product in the product list database, the product is precisely matched via the product identifier. However, if the product in the product knowledge (e.g., Nike Court Vintage Trainers) does not have an exact match in the product list database, the product can be partially matched via the product identifier, where Nike Court Vintage Trainers is matched to the closest related product Nike Air Force 1’07 in the product list database. Matching products in this way can support leveraging product insights from the product knowledge graph.

[0054] The product knowledge graph 120C or the mapped product knowledge graph can be integrated into other systems and components in the product listing system. For example, a product-to-product API 140C can be generated such that product listing system services (e.g., product recommendation applications, product search applications, and seller feedback services) can integrate the mapped product knowledge graph into their systems and use the product-to-product API 140C to provide corresponding services. The product knowledge graph 120C or the mapped version of the product knowledge graph can be stored in the graph database 150C, which supports storing the product knowledge graph 150 in the product listing system and using the product knowledge graph 150.

[0055] Reference Figure 1D , Figure 1D illustrates a schematic diagram associated with providing a knowledge graph using a generative AI KG engine according to an embodiment described herein. Figure 1D illustrates a seed product 102D, a seed product ID 104, a product-to-product API 110D, a product-to-instance API 120D, a ranker 130D, an interface engine 1442D, a first interface 142D, and a second interface 144D. For example, the seed product 102D can be received at the product recommendation service 100D. The seed product 102D is processed to identify the seed product ID 104D of the seed product. The seed product ID 104D is an identifier of the seed product in the product knowledge graph. The product recommendation service 100D can access one or more feature stores associated with the product knowledge graph. For example, the product-to-product API 110D can support mapping the seed product ID 104D to a product in the product listing system database. As discussed, the seed product ID 104D in the product knowledge graph can be mapped to the product ID of an exact match product or the product ID of a non-exact match product in the product listing system database. The product-to-instance API can support mapping the seed product ID 104D to an instance of one or more products (e.g., multiple recommended products) associated with the product ID of a product in the product listing system database.

[0056] The ranker 130D can be used to process multiple recommended products. The ranker 130D can implement a ranking function for ranking multiple recommended products. The multiple recommended products are transmitted to the interface engine 140D, which supports transmitting multiple recommended products for presentation on different types of interfaces. The interface engine 140 can be specifically operated to access additional insights from the LLL to additionally provide multiple recommended products. The first interface 142D can include an instance list 142D_1 (i.e., multiple recommended products), which are instances of products corresponding to the seed product. The first interface 142D also includes edge 1 and edge 2, and edge 1 and edge 2 include edge-related information (e.g., product insights) from the product knowledge graph. The second interface can include an instance list 144D_1 (i.e., multiple recommended products), which are instances of products corresponding to the seed product. The embodiments described herein contemplate other variations and combinations of interfaces associated with providing recommended products and additional insights from the LLM.

[0057] Reference Figure 1E , Figure 1E FIG. shows a schematic diagram associated with an interface that uses a generative AI KG engine in an item listing system to provide product knowledge graph support. The interface can be the product recommendation interface 110E of the buyer, where product recommendations are generated using the product knowledge graph and functions described herein. The product recommendation interface 110E can include product 120E, product 122E, and corresponding product highlight information 124E and product highlight information 126E. The product highlight information can be generated via a generative AI model based on the edge information associated with each product.

[0058] Reference Figure 1F , Figure 1F FIG. shows a schematic diagram associated with an interface that uses a generative AI KG engine in an item listing system to provide product knowledge graph support. The interface can be the product recommendation interface 110F of the buyer, which can be specifically associated with the shopping cart. Based on the shopping cart product 120F, recommended products (i.e., recommended product 122F and recommended product 124F) are generated. The shopping cart product 120F can be a seed product, which is processed using the product knowledge graph to generate recommended products 122F and 124F. The embodiments of the present disclosure contemplate other variations and combinations of interfaces with product knowledge graph support and recommended products.

[0059] Aspects of the technical solution can be described by way of example and with reference to Figure 2A and Figure 2B to describe aspects of the technical solution. Figure 2A is based on reference Figure 6 、 Figure 7 and Figure 8Block diagram of an exemplary technical solution environment for an exemplary environment of an embodiment for implementing a technical solution. Generally, a technical solution environment includes a technical solution system suitable for providing an example item listing system 100 that can adopt the methods of the present disclosure. Specifically, Figure 2A shows a high-level architecture of an item listing system 100 according to an embodiment of the present disclosure. In addition to other engines, managers, generators, selectors, or components not shown (collectively referred to herein as "components"), Figure 2A the item listing platform system 100 corresponds to Figure 1A and Figure 1B .

[0060] Referring to Figure 2A , Figure 2A shows the item listing system 100, the artificial intelligence system 100A, the generative AI KG engine 110, the product knowledge graph generation data 110D, the knowledge graph engine operation 112, the knowledge graph data structure 114, the edge-node prediction service 116, the knowledge graph mapping operation and rules 118, the product knowledge graph 120, the mapped product knowledge graph 122, the item listing system client 130, the machine learning engine 140 including the generative AI model 142, and the item listing system service 110B including the product listing system service 112B and the knowledge graph integration API 110C.

[0061] The generative AI KG engine 110 is responsible for generating and deploying a product knowledge graph (e.g., the product knowledge graph 120 or the mapped product knowledge graph 122). The generative AI KG engine 100 accesses the product knowledge graph generation data 110D associated with generating a product knowledge graph for a product listing system (i.e., the item listing system 100). The product knowledge graph generation data 110D includes a plurality of seed prompt inputs that support the generation of the knowledge graph. By using the generative artificial intelligence (AI) knowledge graph service (i.e., the edge-node prediction service 116), the generative AI KG engine 110 generates a product knowledge graph that includes a plurality of products as nodes and a plurality of relationships as edges. The generative AI knowledge graph service (i.e., the edge-node prediction service 116) is associated with the generative AI model.

[0062] The product knowledge graph 120 can be the mapped product knowledge graph 122. Based on the product identifiers associated with the nodes in the product knowledge graph and the product identifiers associated with a plurality of product instances in the product listing database, the product knowledge graph 120 is mapped to the plurality of product instances in the product listing database. In an example product knowledge graph 120, a first product and a second product in the product knowledge graph are mapped to a first product instance in the product listing database, where the first product is the same as the first product instance, and where the second product is different from the first product instance.

[0063] The product knowledge graph 120 includes multiple products as nodes and multiple relationships as edges. The product knowledge graph 120 is associated with a generative AI knowledge graph service, which is an edge-node prediction service associated with a generative AI model. The nodes in the product knowledge graph include multiple node attributes, and the edges in the product knowledge graph include multiple edge attributes, where the multiple node attributes include a generative AI graph context that includes insights into a first node connected to a second node. Additionally, the nodes in the product knowledge graph 120 include multiple node attributes that include an audience attribute that identifies one or more target demographics of a corresponding product associated with the node.

[0064] The generative AI KG engine 110 generates a mapped product knowledge graph 122 based on mapping multiple products in the product knowledge graph 120 to multiple product instances in the product list database of the product list system 100. The generative AI KG engine 110 generates the mapped product knowledge graph 122 based on multiple mapping rules that support mapping multiple products of the knowledge graph to multiple product instances. A first mapping rule supports matching a first product in the product knowledge graph to a first product instance in the product list database, where the first product is different from the first product instance.

[0065] The generative AI KG engine 110 deploys the product knowledge graph 120 or the mapped product knowledge graph 122 to support one or more services in the product list system 100 (e.g., the item list system service 110B). The product knowledge graph 120 or the mapped product knowledge graph 122 is integrated via an application programming interface with one or more of the following: a product recommendation application, a product search application, and a seller feedback service.

[0066] The product knowledge graph generation data 110D is based on a prompt template associated with multiple seed prompt inputs, where the prompt template includes the following: a seed product element, a requested product quantity element, an audience demographic element, a node element, an edge element, and a recommendation element. The multiple seed prompt inputs are associated with multiple node attributes and multiple edge attributes, and the multiple seed prompt inputs can be executed in batch mode to support generating the product knowledge graph.

[0067] A product listing system service (e.g., product listing system service 112B) in a product listing system accesses a seed product. The product listing system service uses a product knowledge graph to identify a plurality of candidate products associated with the seed product. The plurality of candidate products are individual instances of products associated with corresponding product identifiers. The product instances have a plurality of product features of the instance of the product. By using a ranker (e.g., ranker 114B) of the product listing system, the product listing system service identifies a plurality of recommended products. The plurality of recommended products are a subset of the plurality of candidate products. The product listing system service transmits the plurality of recommended items to generate the plurality of recommended items on a graphical user interface.

[0068] The product listing system service can identify the plurality of candidate products in the following manner: accessing a first product identifier associated with a seed product item and using the first product identifier to query the product knowledge graph. The product listing system service then identifies a plurality of connected product identifiers associated with the first product identifier in the product knowledge graph, where the plurality of connected product identifiers are nodes in the product knowledge graph that are connected to the node of the first product identifier. For each of the first product identifier and the plurality of connected product identifiers, the product listing system service identifies a corresponding candidate product from a product listing database. The product listing system service then identifies the plurality of candidate products based on the corresponding candidate products from the product listing database.

[0069] The product listing system service accesses a product search query. Using a generative artificial intelligence (AI) knowledge graph service, the product listing system service processes the product search query using a product knowledge graph including a plurality of products as nodes and a plurality of relationships as edges, where the product knowledge graph is a mapped product knowledge graph that is mapped to a plurality of product instances in a product listing database. Based on processing the product search query, the product listing system service identifies product search query results from the plurality of product instances; and causes the product search query results to be displayed on a graphical user interface.

[0070] In one embodiment, the product search query can be associated with a seller of the product in the product search query, where processing the product search query is based on an audience attribute associated with the seller of the product in the product knowledge graph and an audience attribute associated with the plurality of products via corresponding nodes of the plurality of products.

[0071] Processing the product search query includes: identifying a first product identifier for the product search query; using the first product identifier to query the product knowledge graph; and identifying a plurality of connected product identifiers associated with the first product identifier in the product knowledge graph, where the plurality of connected product identifiers are nodes in the product knowledge graph that are connected to the node of the first product identifier.

[0072] For each of the first product identifier and the plurality of connected product identifiers, the product list system service identifies corresponding candidate products from a product list database; and identifies product search query results based on the corresponding candidate products from the product list database.

[0073] A product list system client (e.g., item list system client 130) transmits a product search query. Based on the transmitted product search query, product search query results are received, where a generative AI knowledge graph service (which processes the product search query by using a product knowledge graph) is used to generate the product search query results. The product knowledge graph 120 can be a mapped product knowledge graph (e.g., mapped product knowledge graph 122). The product knowledge graph is mapped to a plurality of product instances in the product list database of the product list system. The product list system client causes the product search query results to be displayed on a graphical user interface.

[0074] Reference Figure 2B , Figure 2B shows a generative AI KG engine 110, an item list system client, and an item list system service 140 for providing product knowledge graph functionality. At block 10, the generative AI KG engine 110 accesses product knowledge graph generation data associated with generating a product knowledge graph; at block 12, generates a product knowledge graph that includes a plurality of products as nodes and a plurality of relationships as edges; at block 14, generates a mapped product knowledge graph based on mapping the plurality of products in the product knowledge graph to a plurality of product instances in the product list database of the product list system; and at block 16, deploys the mapped product knowledge graph to support one or more applications in the product list system.

[0075] At block 18, the item list system client 130 transmits a product search query. At block 20, the generative AI KG engine accesses the product search query; at block 24, processes the product search query by using the product knowledge graph mapped to a plurality of product instances in the product list database; at block 26, identifies product search query results from the plurality of product instances; and at block 28, transmits the product search results to the item list system client. At block 34, the item list system service accesses seed products in the product list system; at block 36, identifies a plurality of candidate products associated with the seed products; at block 38, identifies a plurality of recommended products; and at block 40, transmits the plurality of recommended products.

[0076] Example method

[0077] Reference Figure 3 , Figure 4 and Figure 5Flowchart showing a method for providing a product knowledge graph function in an item listing system. The method can be performed using the item listing platform system described herein. In an embodiment, one or more computer storage media embodying computer-executable or computer-usable instructions that, when executed by one or more processors, can cause the one or more processors to perform a method (e.g., a computer-implemented method) in an item listing platform system (e.g., a computerized system or a computer system).

[0078] Go to Figure 3 , a flowchart is provided that shows a method 300 for providing a product knowledge graph function in an item listing system. At block 302, an item listing system with a seed product is provided. At block 304, a generative AI KG engine accesses the seed product. At block 306, the generative AI KG engine uses a product knowledge graph to identify a plurality of candidate products associated with the seed product. The product knowledge graph includes a plurality of products as nodes and a plurality of relationships as edges. At block 308, the generative AI KG engine uses a ranker of the item listing system to identify a plurality of recommended products, where the plurality of recommended products is a subset of the plurality of candidate products. At block 310, the generative AI KG engine transmits the plurality of recommended products.

[0079] Go to Figure 4 , a flowchart is provided that shows a method 400 for providing a product knowledge graph function in an item listing system. At block 402, a generative AI KG engine accesses product knowledge graph generation data associated with generating a product knowledge graph. The product knowledge graph generation data includes a plurality of seed inputs that support the generation of the product knowledge graph. At block 404, the generative AI KG engine uses a generative AI knowledge graph service to generate a product knowledge graph that includes a plurality of products as nodes and a plurality of relationships as edges. The generative AI knowledge graph service is an edge-node prediction service associated with a generative AI model. At block 406, the generative AI KG engine generates a mapped product knowledge graph based on mapping the plurality of products in the product knowledge graph to a plurality of product instances in a product database of the item listing system. At block 408, the generative AI KG engine deploys the mapped product knowledge graph to support one or more services in the item listing system.

[0080] Go to Figure 5, a flowchart is provided which illustrates a method 500 for providing a product knowledge graph function in a project list system. At block 502, a generative AI KG engine accesses a product search query. At block 504, the generative AI KG engine processes the product search query using a product knowledge graph mapped to multiple product instances in a product list database. At block 506, the generative AI KG engine identifies product search query results from the multiple product instances. At block 508, the generative AI KG engine causes the product search results to be displayed on a graphical user interface.

[0081] Technical improvements

[0082] Embodiments of the present invention have been described with reference to several inventive features associated with a project list platform system (e.g., operations, systems, engines, and components). The described inventive features include: operations, interfaces, data structures, and arrangements of computing resources associated with providing the functions described herein with respect to artificial intelligence systems.

[0083] Embodiments of the present invention relate to the field of computing and, more particularly, to a project list system. The exemplary embodiments described below provide a system, method, and program particularly for performing operations of a knowledge graph engine that provide functions associated with generating, deploying, and using a generative AI-based product knowledge graph. Thus, this embodiment improves the technical field of project list system technology by providing more efficient operations and interfaces. For example, the operations and interfaces described for this technical solution provide specific improvements to existing systems, thereby improving the functions of the project list system. Specifically, a product knowledge graph generated using a generative AI model is different from a conventional product knowledge graph. This technical solution solves the problem of the lack of integration of an artificial intelligence system and a generative AI knowledge graph engine in a conventional project list system to improve the operations of the project list system service by providing project list system services based on the use of the product knowledge graph, thereby improving project list system technology.

[0084] The functions of the embodiments of the present invention have been further described by way of embodiments and anecdotal examples - to demonstrate that the operations for providing a product knowledge graph function in a project list system are a solution to a specific problem in project list system technology to improve computing operations and interfaces in the project list system. Additional support for the detailed description of the present invention

[0085] Example project list system environment

[0086] Now refer to Figure 6 , Figure 6 which shows a computing environment of an example project list system 600 in which embodiments of the present disclosure may be employed. Specifically, Figure 6Shows the high-level architecture of an example project listing platform 610 that can host a technology solution environment or a portion thereof. It should be understood that such arrangements and other arrangements described herein are presented as examples. For example, as described above, many of the elements described herein can be implemented as discrete or distributed components or in combination with other components and implemented in any suitable combination and location. Other arrangements and elements (e.g., machines, interfaces, functions, command and function groupings) can be used in addition to or in place of the arrangements and elements shown.

[0087] The project listing system 600 can be a cloud computing environment that provides computing resources for functions associated with the project listing platform 610. For example, the project listing system 600 supports the delivery of computing components and services, including servers, memories, databases, networks, applications, and machine learning associated with the project listing platform 610 and client devices 620. Multiple client devices (e.g., client device 620) include hardware or software for accessing resources on the project listing system 600. The client device 620 can include an application (e.g., client application 622) and interface data (e.g., client application interface data 624) that support client functions associated with the project listing system. Multiple client devices can access the computing components of the project listing system 600 via a network (e.g., network 630) to perform computing operations.

[0088] The project listing platform 610 is responsible for providing a computing environment or architecture that includes infrastructure to support the provision of project listing platform functions (e.g., e-commerce functions). The project listing platform supports storing projects in a project database and provides a search system for receiving queries and identifying search results based on those queries. The project listing platform can also provide a computing environment that has features for managing, selling, buying, and recommending different types of projects. The project listing platform 610 can be dedicated to a content platform, such as the EBAY content platform or e-commerce platform developed by EBAY Inc. in San Jose, California.

[0089] The project listing platform 610 can provide project listing operations 630 and a project listing interface 640. The project listing operations 630 can include service operations, communication operations, resource management operations, security operations, and fault tolerance operations that support specific tasks or functions in the project listing platform 610. The project listing interface 640 can include service interfaces, communication interfaces, resource interfaces, security interfaces, and management and monitoring interfaces that support the functions between project listing platform components. The project listing operations 630 and the project listing interface 640 can enable the communication, coordination, and seamless operation of the project listing system 600.

[0090] As an example, the functions associated with the item listing platform 610 may include shopping operations (e.g., product search and browsing, product selection and cart, checkout and payment, and order tracking); user account operations (e.g., user registration and authentication, and user profile); seller and product management operations (e.g., seller registration and product listing and inventory management); payment and financial operations (e.g., payment processing, refunds and returns); order fulfillment operations (e.g., order processing and fulfillment and inventory management); customer support and communication interfaces (e.g., customer support chat / email and notifications); security and privacy interfaces (e.g., authentication and authorization, payment security); recommendation and personalization interfaces (e.g., product recommendations and customer reviews and ratings); analytics and reporting interfaces (e.g., sales and inventory reports and user behavior analytics); and API and integration interfaces (e.g., APIs for third-party integration).

[0091] The item listing platform 610 may provide an item listing platform database (e.g., item listing platform database 650) to effectively manage and store different types of data. The item listing platform database 650 may include a relational database, a NoSQL database, a search database, a cache database, a content management system, an analytics database, a payment gateway database, a customer relationship management database, a log and error database, an inventory and supply chain database, and a multi-channel database, which are used in combination to effectively manage data and provide an e-commerce experience for users.

[0092] The item listing platform 610 supports applications (e.g., application 660), which are computer programs or software components or services that serve a specific function or set of functions to meet specific item listing platform requirements or user requirements. The applications can be client-side (user-facing) and server-side (backend). The applications can also include applications without any AI support (e.g., application 662), applications supported by traditional AI models (e.g., application 664), and applications supported by generative AI models (e.g., application 666). As an example, the applications can include online storefront applications, mobile shopping applications, implementation and management consoles, payment gateway integrations, user account and authentication applications, search and recommendation engines, inventory and stock management applications, order processing and fulfillment applications, customer support and communication tools, content management systems, analytics and reporting applications, marketing and promotion applications, multi-channel integration applications, log and error tracking applications, customer relationship management (CRM) applications, security applications, and API and web services, which are used in combination to effectively provide an e-commerce experience for users.

[0093] The project list platform 610 may include a machine learning engine (e.g., machine learning engine 670). The machine learning engine 670 refers to a machine learning framework or machine learning platform that provides the infrastructure and tools for designing, training, evaluating, and deploying machine learning models. The machine learning engine 670 can serve as the backbone for developing and deploying machine learning applications and solutions. The machine learning engine 670 can also provide tools for visualizing data and model results and for interpreting model decisions to understand how the model makes predictions.

[0094] The machine learning engine 670 can provide the necessary libraries, algorithms, and utilities to perform various tasks within a machine learning workflow. The machine learning workflow can include data processing, model selection, model training, model evaluation, hyperparameter tuning, scalability, model deployment, inference, integration, customization, data visualization. The machine learning engine 670 can include pre-trained models for various tasks, thus simplifying the development process. In this way, the machine learning engine 670 can simplify the entire machine learning process, from data preparation and model training to deployment and inference, making it accessible and efficiently usable by different types of users (e.g., customers, data scientists, machine learning engineers, and developers) working on a wide range of machine learning applications.

[0095] The machine learning engine 670 can be implemented as a component within the project list system 600 that leverages machine learning algorithms and techniques (e.g., machine learning algorithm 672) to enhance various aspects of the functionality of the project list system. The machine learning engine 670 can provide a range of machine learning algorithms and techniques for teaching a computer to learn from data and make predictions or decisions without explicit programming. These techniques are widely used in various applications across different industries and can include the following examples: supervised learning (e.g., linear regression: classification, support vector machine (SVM)); unsupervised learning (e.g., clustering, principal component analysis (PCA), association rules (e.g., Apriori)); reinforcement learning (e.g., Q-learning, deep Q-network (DQN)); and deep learning (e.g., neural networks, convolutional neural networks (CNN), and recurrent neural networks (RNN)); and ensemble learning random forest.

[0096] Machine learning training data 120 supports the process of building, training, and fine-tuning machine learning models. Machine learning training data 120 consists of a labeled dataset that is used to teach a machine learning model to identify patterns, make predictions, or perform a specific task. Training data typically includes two main components: input features (X) and labels or target values (Y). Input features can include variables, attributes, or characteristics that are used as inputs to the machine learning model. The input features (X) can be numerical, categorical, or even textual, depending on the nature of the problem. For example, in a model used to predict housing prices, the input features might include the number of bedrooms, square footage, neighborhood, etc. Labels or target values (Y) include the values that the model is designed to predict or classify. The labels represent the expected output or ground truth for each corresponding set of input features. For example, in a spam classifier, the label will indicate whether each email is spam (i.e., binary classification). The training process involves presenting the training data to the model, and the model learns to make predictions or decisions by identifying patterns and relationships between the input features (X) and the target values (Y). Machine learning algorithms adjust their internal parameters during training to minimize the difference between their predictions and the actual labels in the training data. The machine learning engine 670 can use historical and real-time data to train models and make predictions, thus continuously improving performance and the user experience.

[0097] The machine learning engine 670 can include machine learning models (e.g., machine learning model 676) generated using the machine learning engine workflow. The machine learning model 676 can include generative AI models and traditional AI models, both of which can be used in the project list system 600. Generative AI models are designed to generate new data (usually in the form of text, images, or other media) based on patterns and knowledge learned from existing data. Generative AI models can be used in various ways, including: content generation, product image generation, personalized product recommendations, natural language chatbots, and content summarization. Traditional AI models cover a wide range of algorithms and technologies and can be used in various ways, including: recommendation systems, predictive analytics, search algorithms, fraud detection, customer segmentation, image classification, natural language processing (NLP), and A / B testing and optimization. In many cases, a combination of generative AI models and traditional AI models can be used to provide a comprehensive and efficient e-commerce experience by combining data-driven insights and creativity.

[0098] A machine learning engine 670 can be used to analyze data, make predictions, and automate processes to provide users with a more personalized and efficient shopping experience. For example, product recommendation search and filtering; pricing optimization, inventory, and stock management; customer segmentation, customer churn prediction and retention, fraud detection, sentiment analysis, customer support and chatbots, image and video analysis, and advertising targeting and marketing. The specific applications of machine learning within the item listing platform 610 can vary depending on the specific goals, available data, and resources.

[0099] Example Distributed Computing System Environment

[0100] Now refer to Figure 7 , Figure 7 , which shows an example distributed computing environment 700 that can adopt the embodiments of the present disclosure. Specifically, Figure 7 , which shows a high-level architecture of an example cloud computing platform 710 that can host a technical solution environment or a part thereof (e.g., a data trustee environment). It should be understood that such arrangements and other arrangements described herein are presented only as examples. For example, as described above, many of the elements described herein can be implemented as discrete or distributed components or in combination with other components and implemented in any suitable combination and location. Other arrangements and elements (e.g., machines, interfaces, functions, command, and function groupings) can be used in addition to or in place of the shown arrangements and elements.

[0101] The data center can support a distributed computing environment 700, which includes a cloud computing platform 710, racks 720, and nodes 730 (e.g., computing devices, processing units, or blades) in the racks 720. A technical solution environment can be implemented using a cloud computing platform 710 that runs cloud services across different data centers and geographical regions. The cloud computing platform 710 can implement a structure controller 740 component for providing and managing resource allocation, deployment, upgrade, and management of cloud services. Generally, the cloud computing platform 710 is used to store data or run service applications in a distributed manner. The cloud computing infrastructure 710 in the data center can be configured to host and support the operation of endpoints of specific service applications. The cloud computing infrastructure 710 can be a public cloud, a private cloud, or a dedicated cloud.

[0102] Node 730 may be provided with a host 750 (e.g., an operating system or a runtime environment) that runs a defined software stack on node 730. Node 730 may also be configured to perform specialized functions (e.g., a compute node or a storage node) within cloud computing platform 710. Node 730 is allocated to run one or more parts of a tenant's service application. A tenant may refer to a customer who utilizes the resources of cloud computing platform 710. The service application components of cloud computing platform 710 that support a particular tenant may be referred to as a multi-tenant infrastructure or a lease. In this document, the terms service application, application, or service may be used interchangeably and broadly refer to any software or part of software that runs on top of a data center or accesses storage devices and compute device locations within a data center.

[0103] When node 730 is supporting multiple individual service applications, node 730 may be partitioned into virtual machines (e.g., virtual machines 752 and 754). A physical machine may also run individual service applications simultaneously. A virtual machine or an entity machine may be configured as a personalized computing environment supported by resources 760 (e.g., hardware resources and software resources) within cloud computing platform 710. It is envisioned that resources may be configured for a particular service application. Additionally, each service application may be divided into functional parts such that each functional part is capable of running on a separate virtual machine. In cloud computing platform 710, multiple servers may be used to run service applications and perform data storage operations in a cluster. Specifically, the servers may perform data operations independently but are exposed as a single device called a cluster. Each server in the cluster may be implemented as a node.

[0104] Client device 780 may be linked to a service application within cloud computing platform 710. Client device 780 may be any type of computing device capable of corresponding to the computing device 700 described in reference Figure 7 above. For example, client device 780 may be configured to issue commands to cloud computing platform 710. In an embodiment, client device 780 may communicate with the service application by directing a communication request to a virtual Internet Protocol (IP) and a load balancer or other means at a specified endpoint within cloud computing platform 710. The components of cloud computing platform 710 may communicate with each other via a network (not shown), which may include but is not limited to one or more local area networks (LANs) and / or wide area networks (WANs).

[0105] Example Computing Environment

[0106] An overview of embodiments of the present invention has been briefly described. The following describes an example operating environment in which embodiments of the present invention may be implemented to provide a general context for aspects of the present invention. Specifically, first refer to Figure 8, an example operating environment for implementing an embodiment of the present invention is shown and generally designated as computing device 800. Computing device 800 is only one example of a suitable computing environment and is not intended to imply any limitation as to the scope of use or functionality of the present invention. Nor should computing device 800 be construed as having any dependency or requirement related to any one or combination of the illustrated components.

[0107] The present invention may be described in the general context of computer code or machine - usable instructions, including computer - executable instructions (such as program modules) executed by a computer or other machine (e.g., a personal data assistant or other handheld device). Generally, program modules including routines, programs, objects, components, data structures, etc., refer to code that performs particular tasks or implements particular abstract data types. The present invention may be practiced in a variety of system configurations including handheld devices, consumer electronics, general - purpose computers, more specialized computing devices, etc. The present invention may also be practiced in a distributed computing environment where tasks are performed by remote processing devices linked through a communication network.

[0108] Referring Figure 8 , computing device 800 includes a bus 810 that directly or indirectly couples the following devices: a memory 812, one or more processors 814, one or more presentation components 816, input / output ports 818, input / output components 820, and an illustrative power supply 822. Bus 810 represents one or more buses (e.g., an address bus, a data bus, or a combination thereof). For the sake of conceptual clarity, Figure 8 the various boxes are shown with lines, and other arrangements of the described components and / or component functions are also contemplated. For example, a presentation component such as a display device may be considered an I / O component. Additionally, a processor has a memory. We recognize this as being in the nature of the art and reiterate Figure 8 that the figures only illustrate example computing devices that may be used in conjunction with one or more embodiments of the present invention. There is no distinction among categories such as "workstation", "server", "laptop computer", "handheld device", etc., because all of these categories are within Figure 8 the scope of and are considered with reference to "computing device".

[0109] Computing device 800 generally includes a variety of computer - readable media. Computer - readable media can be any available media that can be accessed by computing device 800 and includes both volatile and non - volatile media, removable and non - removable media. By way of example and not limitation, computer - readable media may include computer - storage media and communication media.

[0110] Computer storage media include volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information such as computer-readable instructions, data structures, program modules, or other data. Computer storage media include, but are not limited to: RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disks (DVDs) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and that can be accessed by computing device 800. Computer storage media by themselves do not include signals.

[0111] Communication media typically embody computer-readable instructions, data structures, program modules, or other data in a modulated data signal, such as a carrier wave or other transmission mechanism, and include any information delivery media. The term "modulated data signal" refers to a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. By way of example, and not limitation, communication media include wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, and other wireless media. Combinations of any of the above should also be included within the scope of computer-readable media.

[0112] Memory 812 includes computer storage media in the form of volatile and / or non-volatile memory. The memory can be removable, non-removable, or a combination thereof. Exemplary hardware devices include solid state memories, hard disk drives, optical disk drives, etc. Computing device 800 includes one or more processors that read data from various entities such as memory 812 or I / O component 820. Presentation component 816 presents data indications to a user or other device. Exemplary presentation components include display devices, speakers, printing components, vibrating components, etc.

[0113] I / O port 818 allows computing device 800 to be logically coupled to other devices including I / O component 820, some of which may be built in. Illustrative components include microphones, joysticks, gamepads, satellite dishes, scanners, printers, wireless devices, etc.

[0114] Additional structural and functional features of embodiments of the technical solution

[0115] The various components used herein have been identified, and it should be understood that any number of components and arrangements may be employed to achieve the desired functions within the scope of the present disclosure. For example, for clarity of concepts, the components in the embodiments depicted in the figures are shown by lines. Other arrangements of these and other components may also be implemented. For example, although some components are depicted as single components, many of the elements described herein may be implemented as discrete or distributed components or in combination with other components and in any suitable combination and location. Some elements may be entirely omitted. Additionally, as described below, the various functions performed by one or more entities described herein may be performed by hardware, firmware, and / or software. For example, the various functions may be performed by a processor executing instructions stored in a memory. Thus, other arrangements and elements (e.g., machines, interfaces, functions, command, and function groupings) may be used in addition to or in place of the shown arrangements and elements.

[0116] The embodiments described in the following paragraphs may be combined with one or more of the specifically described alternatives. Specifically, the claimed embodiments may include references to more than one other embodiment in the alternatives. The claimed embodiments may specify additional limitations of the claimed subject matter.

[0117] The subject matter of the embodiments of the present invention has been specifically described herein to meet statutory requirements. However, the present specification itself is not intended to limit the scope of the patent. Instead, the inventors have contemplated that the claimed subject matter may also be embodied in other ways in combination with other existing or future technologies to include different steps or combinations of steps similar to those described in this document. Additionally, although the terms "step" and / or "block" may be used herein to denote different elements of the methods employed, such terms should not be construed as implying any particular order among or between the various steps disclosed herein unless and except when the order of the individual steps is explicitly described.

[0118] For the purposes of the present disclosure, the word "comprising" has the same broad meaning as the word "including," and the word "access" includes "receiving," "referencing," or "retrieving." Additionally, the word "communicate" has the same broad meaning as the words "receive" or "send," which are facilitated by a software- or hardware-based bus, receiver, or transmitter using the communication media described herein. Additionally, unless otherwise stated, words such as "a" and "an" include the plural as well as the singular. Thus, for example, in the presence of one or more features, the constraint of "a feature" is satisfied. Additionally, the term "or" includes conjunctive, disjunctive, and both (a or b thus includes a or b, and a and b).

[0119] For purposes of the foregoing detailed discussion, embodiments of the present invention are described with reference to a distributed computing environment; however, the distributed computing environment described herein is merely exemplary. Components may be configured to perform novel aspects of the embodiments, where the term "configured to" may mean "programmed to" perform a particular task or implement a particular abstract data type using code. Further, while embodiments of the present invention may generally be described with reference to the technical solution environments and schematic diagrams described herein, it should be understood that the described technology may be extended to other implementation contexts.

[0120] Embodiments of the present invention have been described with respect to specific embodiments that are intended in all respects to be illustrative and not restrictive. Alternative embodiments will become apparent to those of ordinary skill in the art to which the present invention pertains without departing from the scope of the present invention.

[0121] As can be seen from the foregoing, the invention is well suited to achieve all the objects and purposes set forth above herein, as well as other obvious and inherent advantages of the structure.

[0122] It should be understood that certain features and subcombinations are useful and may be employed without reference to other features or subcombinations. This is contemplated by the claims and is within the scope of the claims.

Claims

1. A computerized system comprising: one or more computer processors; as well as A computer memory storing computer usable instructions that, when used by the one or more computer processors, cause the one or more computer processors to perform operations comprising: Provide a product listing system with seed products; access to said seed products; using a product knowledge graph to identify a plurality of candidate products associated with the seed product, Wherein, the product knowledge graph is associated with a generative AI knowledge graph service; identifying a plurality of recommended products, wherein the plurality of recommended products is a subset of the plurality of candidate products; and The plurality of recommended products are transmitted so that the plurality of recommended products are generated on a graphical user interface.

2. The system according to claim 1, wherein: The plurality of candidate products are identified based on: accessing a first product identifier associated with a seed product item; Using the first product identifier, querying the product knowledge graph; identifying, in the product knowledge graph, a plurality of connected product identifiers associated with the first product identifier, wherein the plurality of connected product identifiers are nodes in the product knowledge graph connected to a node of the first product identifier; for the first product identifier and each of the plurality of connected product identifiers, identifying a corresponding candidate product from a product list database; and The plurality of candidate products are identified based on the corresponding candidate products from the product listing database.

3. The system according to claim 2, wherein: The plurality of candidate products are respective instances of a product associated with a corresponding product identifier, the instances of the product having a plurality of product features of the instance of the product.

4. The system according to claim 1, wherein: The product knowledge graph includes a plurality of products as nodes and a plurality of relationships as edges, wherein the nodes in the product knowledge graph include a plurality of node attributes, and the edges in the product knowledge graph include a plurality of edge attributes, wherein the plurality of node attributes include a generative AI graph context, the generative AI graph context includes insights into a first node connected to a second node, and the plurality of node attributes include audience attributes, the audience attributes identifying one or more target demographics for a corresponding product associated with the node.

5. The system according to claim 1, wherein: The generative AI knowledge graph service is an edge-node prediction service associated with a generative AI model, and wherein identifying the plurality of candidate products associated with the seed product is performed using a ranker of the product listing system.

6. The system according to claim 1, wherein: The product knowledge graph is a mapped product knowledge graph, and the product knowledge graph is mapped to multiple product instances in a product list database based on product identifiers associated with nodes in the product knowledge graph and product identifiers associated with multiple product instances in a product list database.

7. The system according to claim 6, wherein: A first product and a second product in the product knowledge graph are mapped to a first product instance in the product list database, wherein the first product is the same as the first product instance, and wherein the second product is different from the first product instance.

8. The system of claim 1, wherein the operations further comprise: Accessing product knowledge graph generation data associated with generating a product knowledge graph of the product list system, wherein the product knowledge graph generation data includes a plurality of seed prompt inputs supporting generation of the product knowledge graph; Using the generative artificial intelligence (AI) knowledge graph service, generating the product knowledge graph, wherein the product knowledge graph includes a plurality of products as nodes; and The product knowledge graph is deployed to support one or more applications in the product listing system.

9. The system of claim 1, wherein the operations further comprise: Access product search queries; Using the generative AI knowledge graph service, processing the product search query using the product knowledge graph, wherein the product knowledge graph is a mapped product knowledge graph, and the product knowledge graph is mapped to a plurality of product instances in a product list database; Based on processing the product search query, identifying product search query results from the plurality of product instances; and The product search query results are caused to be displayed on a graphical user interface.

10. The system of claim 1, the operations further comprising: Send product search queries; Based on transmitting the product search query, receiving a product search query result, wherein the product search query result is generated using a generative AI knowledge graph service that processes the product search query by using the product knowledge graph, wherein the product knowledge graph is a mapped product knowledge graph that is mapped to a plurality of product instances in a product list database; and The product search query results are caused to be displayed on a graphical user interface.

11. One or more computer storage media having computer executable instructions thereon, which when executed by a computing system having a processor and a memory, cause the processor to perform operations comprising: Accessing product knowledge graph generation data associated with a product knowledge graph of a system for generating a product list, wherein the product knowledge graph generation data includes a plurality of seed prompt inputs supporting generation of the product knowledge graph; Generate the product knowledge graph by using a generative artificial intelligence (AI) knowledge graph service, wherein the product knowledge graph includes a plurality of products as nodes and a plurality of relationships as edges; generating a mapped product knowledge graph based on mapping the plurality of products in the product knowledge graph to a plurality of product instances in a product listing database of a product listing system; and The mapped product knowledge graph is deployed to support one or more services in the product listing system.

12. The medium according to claim 11, wherein The product knowledge graph generates data based on a prompt template associated with the plurality of seed prompt inputs, wherein the prompt template includes the following: a seed product element, a requested product quantity element, an audience demographic element, a node element, an edge element, and a recommendation element.

13. The medium according to claim 11, wherein The plurality of seed hint inputs are associated with a plurality of node attributes and a plurality of edge attributes, and the plurality of seed hint inputs are executable in a batch mode to support generating the product knowledge graph.

14. The medium according to claim 11, wherein Generating the mapped product knowledge graph is based on multiple mapping rules that support mapping multiple products in the knowledge graph to the multiple product instances, wherein a first mapping rule supports matching a first product in the product knowledge graph to a first product instance in the product list database, and the first product is different from the first product instance.

15. The medium according to claim 11, wherein The mapped product knowledge graph is integrated with each of: a product recommendation application, a product lookup application, and a seller feedback service via an application programming interface.

16. A computer-implemented method, the method comprising: Access product search queries; Using a generative artificial intelligence (AI) knowledge graph service, processing the product search query using a product knowledge graph including a plurality of products as nodes and a plurality of relationships as edges, wherein the product knowledge graph is a mapped product knowledge graph, and the product knowledge graph is mapped to a plurality of product instances in a product list database; Based on processing the product search query, identifying product search query results from the plurality of product instances; and The product search query results are caused to be displayed on a graphical user interface.

17. The method according to claim 16, wherein: Processing the product search query includes: identifying a first product identifier for the product search query; Using the first product identifier, querying the product knowledge graph; and A plurality of connected product identifiers associated with the first product identifier are identified in the product knowledge graph, wherein the plurality of connected product identifiers are nodes in the product knowledge graph connected to a node of the first product identifier.

18. The method of claim 17, wherein the operation further comprises: for the first product identifier and each of the plurality of connected product identifiers, identifying a corresponding candidate product from the product list database; as well as The product search query results are identified based on the corresponding candidate products from the product listing database.

19. The method according to claim 16, wherein: The product knowledge graph is mapped to the plurality of product instances in the product list database based on product identifiers associated with nodes in the product knowledge graph and product identifiers associated with the plurality of product instances in the product list database.

20. The method according to claim 19, wherein: The product search query is associated with a seller of a product to which the product search query corresponds, wherein processing the product search query is based on audience attributes associated with the seller of the product in the product knowledge graph and audience attributes associated with the multiple products via corresponding nodes of the multiple products.

Images