Abstract: Examples provide improved methods for generating search recommendations in response to a user-initiated search. Examples include receiving a search request including search terms; identifying one or more product categories as output from a machine learning classification model; identifying products that are assigned to those product categories, including product titles short descriptions in a natural language; applying the product titles and short descriptions as input to a second machine learning model that is configured to generate recommended searches; scoring each recommended search of the plurality of recommended searches; selecting one or more recommended searches of the plurality of recommended searches based on the scoring; and causing the one or more recommended searches to be displayed as user-interactable components on a graphical user interface, each user-interactable component being configured to execute a second search request upon user interaction with the user-interactable component.

Abstract: Systems and methods for generating an electronic planogram optimized for space elasticity calculating a space elasticity include training a constrained linear regression model using a first training dataset, training a multiple non-linear regression model based on a second training dataset including, training a non-linear multiple-constraint mixed integer optimization model based at least in part on an output of the constrained linear regression model and an output of the multiple non-linear regression model, and generating an electronic planogram optimized for space elasticity. For each item category, a space elasticity is determined by fitting the constrained linear regression model, the multiple non-linear regression model is applied to determine a cross-space elasticity, and the non-linear multiple-constraint mixed integer optimization model generates a category space allocation. The category space allocation for each item category is aggregated to generate the electronic planogram.

Abstract: Example implementations may relate to systems and methods for rendering colors on surfaces while preserving lighting. For example, a computer-implemented method may include receiving an input image. The computer implemented method can also include receiving a binary mask comprising a region of interest. The computer implemented method can further include generating a greyscale image of the input image. The computer-implemented method can additionally include determining a mean pixel value for a region of interest of the greyscale image, wherein the region of interest of the greyscale image is equivalent to the region of interest of the binary mask. The computer-implemented method can also include generating a gamma corrected image of the greyscale image based on the mean pixel value for the region of interest of the greyscale image.

Abstract: A data generation system can include a computing device that is configured to receive a request to generate a training dataset for an attribute and identify a set of item identifiers from an item database based on an engagement indication. The computing device is further configured to, for each item identifier of the set of item identifiers, obtain a query list including queries resulting in an engagement between the corresponding item identifier and a user and, in response to a portion of queries of the query list including the attribute being above a threshold, assign the corresponding item identifier to the training dataset for the attribute. The computing device is also configured to store the training dataset for the attribute in a training dataset database.

Abstract: Some embodiments relate to controlling digital communications. An example implementation may cause an execution of a trained first model applied to historic features of groupings of items and obtaining a set of carousel identifiers; obtain session context corresponding to a first set of items during a current network session; perform a similarity search in the collection of the sets of carousel identifiers, and identify a first set of carousel identifiers having a threshold relationship with the session context; determine a ranking of the first set of carousel identifiers and identify a first carousel identifier having an estimated highest ranked association with the network session; and control the data communications transceiver to transmit carousel content data in controlling the user computing device to render the carousel content data comprising the first carousel identifier consistent with the carousel interface.

Abstract: Example implementations relate to automatically identifying matching data entities. A determination request is received and for each of a first data entity and a second data entity, a sequence of demarcated attributes is generated. A similarity score is generated, using a trained classification model, between the first data entity and the second data entity that accounts for feature interactions between the first data entity and the second data entity. The trained classification model receives the demarcated attributes for each of the first data entity and the second data entity as inputs, and is trained using a dataset of annotated pairs of data entities. Weights in the trained classification model are determined based on a customizable loss function. In accordance with a determination that the similarity score is greater than a predetermined threshold, an indication that the first data entity matches the second data entity is generated.

Abstract: Systems and methods for predicting advertising related inventory and opportunity for advertising campaigns based on metadata and dimensionality reduction are disclosed. In some embodiments, a disclosed method includes: receiving, from a computing device, a prediction request associated with an advertising campaign; determining a metadata set associated with the advertising campaign; generating, based on a machine learning model, an embedding for the metadata set; determining, based on the embedding, at least one embedding associated with at least one advertising opportunity data; generating predicted advertising opportunity data for the advertising campaign based on the at least one advertising opportunity data; and transmitting the predicted advertising opportunity data to the computing device.

Abstract: In some embodiments, apparatuses and methods are provided herein useful to processing captured images. In some embodiments, there is provided a system for processing captured images of objects at a product storage facility including a trained machine learning model; and a control circuit. The control circuit may identify a product identifier associated with an object in a captured image; generate predicted product identifiers associated with the object in the captured image based on text identified from the object in the captured image; aggregate the predicted product identifiers; determine a feature of the objects associated with the aggregated predicted product identifiers; determine one or more confusing product identifiers based on a determination of the aggregated predicted product identifiers being associated with the feature; and update a dataset with at least one of the one or more confusing product identifiers and images associated with the one or more confusing product identifiers.

Abstract: Provided are check-in to checkout systems and methods. The system includes a point of sale, a display, a mobile device, and a backend system. When the POS starts a transaction, such as, but not limited to scanning an item, the POS generates a transaction/basket identification. In response to starting the transaction, the POS sends the transaction/basket ID to the display in the form of a QR code or one-dimensional barcode. The mobile device operating a mobile application scans and captures the displayed code and reports to the backend systems. The backend systems then send a check-in event, which includes gathered data by the backend systems, to the POS and the POS acknowledges receipt of the message. At this point, a notification may be sent to the mobile application that the check-in was successful and a link is established for flow of information.

Abstract: Systems and methods for automatically generating electronic communications including relevant content elements are disclosed. A request to generate an electronic communication including a user identifier and a prior interaction is received. A hybrid sequential model is implemented to generate a set of hybrid sequential element recommendations selected from a catalog of items. The hybrid sequential model includes a sequential sub-model to select a set of sequential elements and a re-ranking sub-model to re-rank the sequential elements to generate the hybrid sequential elements. An affinity model is implemented to generate affinity element recommendations and a set of relevant content elements including at least one of the set of hybrid sequential element recommendations and at least one of the set of affinity element recommendations is generated. The electronic communication is generated including the set of relevant content elements and is transmitted to a user device associated with the user identifier.

Abstract: In some embodiments, apparatuses and methods are provided herein useful to detecting a mis-scan of an item. In some embodiments, there is provided a system for detecting a mis-scan of an item for purchase comprising a checkout station; a first staging location; a second staging location; a first area of interest at the checkout station; a second area of interest at the checkout station; a camera; and a control circuit configured to: receive an identifier of a first item; detect a hand of a user purchasing the first item, the first item, and a scanner on a first image captured by the camera; in response to the detection of the hand, the first item, and the scanner, initiate detection of mis-scan items during a checkout process; determine that a payment transaction has been received; and stop the detection of mis-scan items.

Abstract: A system comprising one or more processors and one or more non-transitory computer-readable media storing computing instructions, that when executed on the one or more processors, cause the one or more processors to perform: receiving stock images comprising an anchor garment; automatically identifying the anchor garment and complementary garments within the stock images; selecting an image of the stock images in which a mask area of a first complementary garment of the complementary garments as a ratio of an area of the anchor garment is largest over other complementary garments of the complementary garments; performing an image search, using the image, in an item catalog for similar garments to the first complementary garment; and displaying, on a user interface, an avatar wearing the anchor garment and at least one of the similar garments. Other embodiments are disclosed.

Abstract: Examples provide conversational item search refinement using multi-facet filtering of items in a catalog. A multi-facet filter manager extracts facets and actions corresponding to the facets from a user utterance. The facet-actions include an entity-role and one or more filter actions associated with the facets. A facet-action includes filter actions such as exact, exclude, greater than, less than, etc. Multi-facet filters corresponding to the facet-actions are applied to a plurality of items in the catalog. The candidate items remaining after filtering are scored. The scores indicate relevance of each candidate item. One or more of the candidate items with the highest scores are selected. The selected items are added to search results which are returned to the user in response to the conversational search query. The multi-facet filter manager enables faster and more accurate search results using fewer conversational turns for reduced system resource usage and improved user efficiency.

Abstract: Systems and methods for hybrid optimization of training ranking models is disclosed. A training dataset including a plurality of anchor items, a plurality of recommended item sets, and ground truth data is obtained from a database. A base machine learning model including a step function configured to determine a relevance score is iteratively trained to generate a trained ranking model. The plurality of anchor items and the plurality of recommended item sets are provided as an input to the base machine learning model and the ground truth is provided as a target output. The step function is trained using an adaptive step size according to a first order Barzilai-Borwein (BB) method and a line search method. The trained ranking model is stored in non-transitory memory.