# US 2021/0241350 A1 · Gender attribute assignment using a multimodal neural graph > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2021/0241350 A1 - Document ID: US20210241350A1 - Status: Published application - Kind: A1 - Application number: 16/779,545 - Filed: 2020-01-31 - Published: 2021-08-05 - Earliest claimed: 2020-01-31 - Assignees: Walmart Apollo, LLC - Applicants: Walmart Apollo, LLC - Inventors: Mansi Ranjit Mane; Anirudha Sundaresan; Aditya Mantha; Stephen Dean Guo; Kannan Achan - Web page: https://patentgrep.com/patent/US20210241350A1 ## CPC - G06Q 30/0631: {Recommending goods or services} - G06F 16/9024: {Graphs; Linked lists (G06F16/9027 takes precedence)} - G06F 16/906: Clustering; Classification - G06F 40/20: Natural language analysis (semantic analysis of natural language G06F40/30) - G06F 40/30: Semantic analysis - G06N 3/045: Combinations of networks - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06N 3/048: Activation functions - G06N 3/08: Learning methods - G06N 3/0895: Weakly supervised learning, e.g. semi-supervised or self-supervised learning - G06N 3/09: Supervised learning - G06Q 30/0603: {Catalogue creation or management} - G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract A system including one or more processors and one or more non-transitory computer-readable media storing computing instructions configured to run on the one or more processors and perform receiving from an item catalog database a respective item description and respective attribute values for each item of a set of items; generating text embeddings using a text embedding model to represent the respective item description and the respective attribute values; generating a graph of the set of items from the item catalog database connected by a set of edges; training the text embedding model and a machine learning model using a neural loss function based on the graph; and automatically determining, based on the machine learning model, as trained, a gender label for each first item in which the gender classification is unlabeled and in which a respective quantity of respective attribute values for the each first item is at least a predetermined threshold. Other embodiments are disclosed. ## Claims ### Claim 1 A system comprising: one or more processors; and one or more non-transitory computer-readable media storing computing instructions configured to run on the one or more processors and perform: receiving from an item catalog database a respective item description and respective attribute values for each item of a set of items, wherein a gender classification of the respective attribute values for the each item of the set of items is either labeled or unlabeled; generating text embeddings using a text embedding model to represent the respective item description and the respective attribute values for the each item of the set of items; generating a graph of the set of items from the item catalog database connected by a set of edges, wherein each pair of items of the set of items that is connected by a respective edge of the set of edges in the graph has been viewed together in one or more respective sessions, the respective edge comprises a weight comprising a co-view count, and the set of edges comprises (a) one or more unlabeled-unlabeled edges, (b) one or more labeled-unlabeled edges, and (c) one or more labeled-labeled edges; training the text embedding model and a machine learning model using a neural loss function based on the graph; and automatically determining, based on the machine learning model, as trained, a gender label for each first item of the set of items in which the gender classification is unlabeled and in which a respective quantity of respective attribute values for the each first item is at least a predetermined threshold. ### Claim 2 (depends on 1) The system of claim 1, wherein the computing instructions are further configured to perform: determining, based on an image embedding model, as trained, a gender label for each second item of the set of items that does not meet the predetermined threshold. ### Claim 3 (depends on 1) The system of claim 1, wherein the predetermined threshold is 5. ### Claim 4 (depends on 1) The system of claim 1, wherein the computing instructions are further configured to perform: transforming an image into a second vector representing the image using a residual neural network (“ResNet”). ### Claim 5 (depends on 1) The system of claim 1, wherein the computing instructions are further configured to perform: training an image embedding model based on images of items from the item catalog database using loss equations to minimize a distance between text representations and image representations for the items. ### Claim 6 (depends on 5) The system of claim 5, wherein the images depict items of clothing from the item catalog database. ### Claim 7 (depends on 1) The system of claim 1, wherein: the text embedding model is a Bidirectional Encoder Representations from Transformers (“BERT”); and an output from the text embedding model comprises a vector representation. ### Claim 8 (depends on 1) The system of claim 1, wherein training the text embedding model and the machine learning model using the neural loss function based on the graph further comprises: training the machine learning model with the neural loss function based on first distances between first text embeddings for first pairs of nodes connected by the one or more labeled-labeled edges, second distances between second text embeddings for second pairs of nodes connected by the one or more labeled-unlabeled edges, third distances between third text embeddings for third pairs of nodes connected by the one or more unlabeled-unlabeled edges, and a softmax loss cost function for fourth text embeddings of nodes of the graph that are labeled. ### Claim 9 (depends on 1) The system of claim 1, wherein the gender classification, when labeled, comprises one of: a male gender label; a female gender label; or a unisex gender label. ### Claim 10 (depends on 1) The system of claim 1, wherein the computing instructions are further configured to perform: receiving a selection of an anchor item from a user, the anchor item comprising a first gender label; determining one or more recommended items that match the first gender label based on the gender labels determined by the machine learning model; and sending instructions to display the one or more recommended items to the user. ### Claim 11 The method being implemented via execution of computing instructions configured to run at one or more processors and stored at one or more non-transitory computer-readable media, the method comprising: receiving from an item catalog database a respective item description and respective attribute values for each item of a set of items, wherein a gender classification of the respective attribute values for the each item of the set of items is either labeled or unlabeled; generating text embeddings using a text embedding model to represent the respective item description and the respective attribute values for the each item of the set of items; generating a graph of the set of items from the item catalog database connected by a set of edges, wherein each pair of items of the set of items that is connected by a respective edge of the set of edges in the graph has been viewed together in one or more respective sessions, the respective edge comprises a weight comprising a co-view count, and the set of edges comprises (a) one or more unlabeled-unlabeled edges, (b) one or more labeled-unlabeled edges, and (c) one or more labeled-labeled edges; training the text embedding model and a machine learning model using a neural loss function based on the graph; and automatically determining, based on the machine learning model, as trained, a gender label for each first item of the set of items in which the gender classification is unlabeled and in which a respective quantity of respective attribute values for the each first item is at least a predetermined threshold. ### Claim 12 (depends on 11) The method of claim 11, further comprising: determining, based on an image embedding model, as trained, a gender label for each second item of the set of items that does not meet the predetermined threshold. ### Claim 13 (depends on 11) The method of claim 11, wherein the predetermined threshold is 5. ### Claim 14 (depends on 11) The method of claim 11, further comprising: transforming an image into a second vector representing the image using a residual neural network (“ResNet”). ### Claim 15 (depends on 11) The method of claim 11, further comprising: training an image embedding model based on images of items from the item catalog database using loss equations to minimize a distance between text representations and image representations for the items. ### Claim 16 (depends on 11) The method of claim 11, wherein the images depict items of clothing from the item catalog database. ### Claim 17 (depends on 11) The method of claim 11, wherein: the text embedding model is a Bidirectional Encoder Representations from Transformers (“BERT”); and an output from the text embedding model comprises a vector representation. ### Claim 18 (depends on 11) The method of claim 11, wherein training the text embedding model and the machine learning model using the neural loss function based on the graph further comprises: training the machine learning model with the neural loss function based on first distances between first text embeddings for first pairs of nodes connected by the one or more labeled-labeled edges, second distances between second text embeddings for second pairs of nodes connected by the one or more labeled-unlabeled edges, third distances between third text embeddings for third pairs of nodes connected by the one or more unlabeled-unlabeled edges, and a softmax loss cost function for fourth text embeddings of nodes of the graph that are labeled. ### Claim 19 (depends on 11) The method of claim 11, wherein the gender classification, when labeled, comprises one of: a male gender label; a female gender label; or a unisex gender label. ### Claim 20 (depends on 11) The method of claim 11, further comprising: receiving a selection of an anchor item from a user, the anchor item comprising a first gender label; determining one or more recommended items that match the first gender label based on the gender labels determined by the machine learning model; and sending instructions to display the one or more recommended items to the user.