Abstract: Deriving insights from time series data can include receiving subject matter expert (SME) input characterizing one or more aspects of a time series. A model template that specifies one or more components of the time series can be generated by translating the SME input using a rule-based translator. A machine learning model based on the model template can be a multilayer neural network having one or more component definition layers, each configured to extract one of the one or more components from time series data input corresponding to an instantiation of the time series. With respect to a decision generated by the machine learning model based on the time series data input, a component-wise contribution of each of the one or more components to the decision can be determined. An output can be generated, the output including the component-wise contribution of at least one of the one or more components.

Abstract: Methods, systems, and computer program products for facilitating user selection using trend-based joint embeddings are provided herein. A method includes obtaining a selection of an item in an online catalog; determining a compatible item of the plurality of items at least in part by providing the selected at least one item and at least one previously selected item corresponding to the user to a trend-based machine learning model, wherein the trend-based machine learning model is trained on historical data associated with the item in the online catalog and fine-tuned based on current trend data from multiple data sources; receiving feedback in response to outputting the at least one compatible item; identifying one or more attributes related to the at least one compatible item based on the feedback; and using the trend-based machine learning model to determine at least one additional compatible item based on the one or more attributes.

Abstract: A method, system, and computer program product for an interpretable, feature-based post-hoc black box explainer for univariate time series forecasters are provided. The method receives a set of time series forecasting predictions. The set of time series forecasting predictions are generated from a set of black-box models trained with an initial data set. The method generates a set of features based on at least a portion of the initial data set. A set of surrogate models are trained based on the set of time series forecasting predictions and at least a portion of the set of features. A subset of surrogate models is selected. Based on the subset of surrogate models, the method generates one or more explanation outputs for time series forecasting predictions of the set of black-box models.

Abstract: One embodiment provides a method, including: receiving information for each of a plurality of products; for each of the plurality of products: predicting a demand for the product; calculating a sustainability score, wherein the sustainability score is based upon a material composition of the product; and computing a similarity score, wherein the computing comprises comparing pairs of products within the plurality of products; generating a plurality of assortment choices from the plurality of products, wherein each of the plurality of assortment choices has a corresponding assortment demand value, assortment sustainability score, and assortment similarity score, wherein the generating comprises utilizing a multi-objective formulation based upon the predicted demand, the calculated sustainability score, and the computed similarity score; and providing the plurality of assortment choices to a user for selection.

Abstract: Methods, systems, and computer program products for determining locations for offerings using artificial intelligence are provided herein. A computer-implemented method includes determining an initial sub-set of one or more enterprise locations, from a set of multiple enterprise locations, for providing at least one enterprise offering; determining at least one sequential ordering of one or more additional ones of the multiple enterprise locations for providing the at least one enterprise offering by processing data related to the initial sub-set of one or more enterprise locations using an artificial intelligence engine that utilizes one or more artificial intelligence techniques; and automatically training at least a portion of the artificial intelligence engine that utilizes the one or more artificial intelligence techniques based at least on feedback data pertaining to provision of the at least one enterprise offering to the at least one sequential ordering of one or more additional enterprise locations.

Abstract: Methods, systems, and computer program products for facilitating user selection using trend-based joint embeddings are provided herein. A method includes obtaining a selection of an item in an online catalog; determining a compatible item of the plurality of items at least in part by providing the selected at least one item and at least one previously selected item corresponding to the user to a trend-based machine learning model, wherein the trend-based machine learning model is trained on historical data associated with the item in the online catalog and fine-tuned based on current trend data from multiple data sources; receiving feedback in response to outputting the at least one compatible item; identifying one or more attributes related to the at least one compatible item based on the feedback; and using the trend-based machine learning model to determine at least one additional compatible item based on the one or more attributes.

Abstract: Methods, systems, and computer program products for customizing agricultural practices to maximize crop yield are provided herein. A computer-implemented method includes obtaining data pertaining to (i) a geographical area comprising a plurality of regions and (ii) one or more agricultural practices applied to the geographical area; assigning each of the plurality of regions to a respective cluster of a set clusters, based at least in part on comparing features identified in the data, wherein similar ones of said regions are assigned to the same cluster; generating instructions that are specific to a given cluster in the set, wherein the instructions relate to agricultural tasks to be performed on the regions assigned to the given cluster; and triggering, based on said instructions, one or more automated farming processing devices, thereby carrying out at least a portion of said agricultural tasks.

Abstract: Methods, systems, and computer program products for environmental impact aware product refurbishing are provided herein. A computer-implemented method includes obtaining information for products comprising images of the products and location-specific demand data; determining product embeddings of the products based on the images, wherein each of the product embeddings encodes attributes of the corresponding product; creating one or more refurbished designs of each given one of the products based on the initial image of the given product and one or more design constraints; calculating an environmental impact score and a demand impact score associated with each of the created refurbished designs, wherein the demand impact score is based on the location-specific demand data; generating a recommendation to refurbish at least one of the products in accordance with at least one of the refurbished designs based on the environmental impact scores and the demand impact scores.

Abstract: Machine logic for selecting a given item for an inventory. This selection of the given item is based, at least in part, upon: (i) an amount of “ensembles” that include the item; (ii) relative popularity of “ensembles” that contain the item; and/or (iii) the relative profitability of ensembles that includes the item. This technology can be provided as part of assortment planning software for retail stores selling items such as: fashionable clothing, furniture sets, jewelry sets, and other types of items that are typically sold in ensembles and have subjective factors (like aesthetics) that play into the attractiveness of the ensemble considered as a whole.

Abstract: One embodiment provides a computer implemented method, including: creating a buyer and manufacturer virtual agent representing a buyer and manufacturer within a design supply chain; creating a designer virtual agent representing a designer within the design supply chain; detecting a conflict between design attributes preferred by the buyer and manufacturer and design attributes preferred by the designer for a product being designed by the designer; resolving, utilizing the buyer and manufacturer virtual agent and the designer virtual agent, the conflict via initiating a conversation between the buyer and manufacturer virtual agent and designer virtual agent to generate attribute suggestions for incorporation into the product; and presenting the attribute suggestions to the designer for incorporation into the product.

Abstract: One embodiment provides a method, including: receiving information for each of a plurality of products; for each of the plurality of products: predicting a demand for the product; calculating a sustainability score, wherein the sustainability score is based upon a material composition of the product; and computing a similarity score, wherein the computing comprises comparing pairs of products within the plurality of products; generating a plurality of assortment choices from the plurality of products, wherein each of the plurality of assortment choices has a corresponding assortment demand value, assortment sustainability score, and assortment similarity score, wherein the generating comprises utilizing a multi-objective formulation based upon the predicted demand, the calculated sustainability score, and the computed similarity score; and providing the plurality of assortment choices to a user for selection.

Abstract: Methods, systems, and computer program products for customizing agricultural practices to maximize crop yield are provided herein. A computer-implemented method includes obtaining data pertaining to (i) a geographical area comprising a plurality of regions and (ii) one or more agricultural practices applied to the geographical area; assigning each of the plurality of regions to a respective cluster of a set clusters, based at least in part on comparing features identified in the data, wherein similar ones of said regions are assigned to the same cluster; generating instructions that are specific to a given cluster in the set, wherein the instructions relate to agricultural tasks to be performed on the regions assigned to the given cluster; and triggering, based on said instructions, one or more automated farming processing devices, thereby carrying out at least a portion of said agricultural tasks.

Abstract: Methods, systems, and computer program products in connection with a system for bidding on multi-modal sub-spaces for e-commerce searches is provided herein. A computer-implemented method includes representing, in an embedding space affiliated with at least one electronic commerce website: (i) images related to one or more bids and (ii) items of text related to the one or more bids, the images and the items of text both coming from one or more bidding entities. Such a method also includes mapping a query to the embedding space based on a comparison of features of the query to (i) the images and (ii) the items of text. Further, such a method additionally includes selecting at least one of the one or more bidding entities based on the mapping, and outputting, to at least one user in response to the query, search results associated with the at least one selected bidding entity.

Abstract: One embodiment provides a method, including: receiving a client fashion catalog comprising a plurality of images of client wearable products; creating a market fashion catalog comprising a plurality of market wearable products identified from sources other than the client fashion catalog, wherein the creating comprises (i) accessing a plurality of secondary sources, (ii) capturing information corresponding to market wearable products from the plurality of secondary sources, and (iii) aggregating the information to create the market fashion catalog; generating, for each market wearable product, (iv) a desirability score and (v) a multi-modal description of characteristics; producing, for the market wearable products, by utilizing the multi-modal descriptions, a similarity score indicating how similar the market wearable product is to products within the client fashion catalog, wherein the producing comprises comparing market wearable products to client wearable products; and providing a recommendation for a ch

Abstract: Methods, systems, and computer program products for generating search determinations for assortment planning and buying using visual sketches are provided herein. A computer-implemented method includes processing a query image by identifying one or more visual features in the query image and applying at least one nearest neighbor algorithm to the one or more identified visual features; identifying, from one or more databases, multiple images based at least in part on the processing; generating a result set by applying one or more smoothing algorithms to the multiple identified images; generating at least one sketch based at least in part on the result set; and outputting the at least one generated sketch to one or more users via a user interface.

Abstract: Methods, systems, and computer program products for generating metadata for image-based querying are provided herein. A computer-implemented method includes processing a query image against a database by applying a deep learning visual model to the query image and images contained within the database; retrieving a set of multiple images from the database based on the processing; identifying subsets of images among the set of multiple images by analyzing metadata attribute values of the set of multiple images and nearest neighbor distance values between the query image and the multiple images in the set; determining one or more items of metadata attributable to the query image by processing metadata of the subsets of images; and outputting, to a user, the items of metadata in response to the query image.

Abstract: A computer scans user data for upcoming events and identifies a potential gift giving event from these events using contextual information about the events to determine the events' nature. The computer identifies a plurality of event attendees and analyzes information about two or more of the event attendees. The computer groups a plurality of people into a group of potential gift collaborators, wherein the plurality of people is chosen from the plurality of attendees. The computer generates a first potential gift list. The computer analyzes gift giving abilities of the group members and selects, based partially upon the analysis, a gift giver. The computer generates a second potential gift list with detailed information about one or more potential gifts wherein the gifts are based partially on the first potential gift list. The computer provides at least a portion of the second potential gift list to the selected gift giver.

Abstract: Methods, systems, and computer program products for generating metadata for image-based querying are provided herein. A computer-implemented method includes processing a query image against a database by applying a deep learning visual model to the query image and images contained within the database; retrieving a set of multiple images from the database based on the processing; identifying subsets of images among the set of multiple images by analyzing metadata attribute values of the set of multiple images and nearest neighbor distance values between the query image and the multiple images in the set; determining one or more items of metadata attributable to the query image by processing metadata of the subsets of images; and outputting, to a user, the items of metadata in response to the query image.

Abstract: Methods, systems, and computer program products for generating search determinations for assortment planning and buying using visual sketches are provided herein. A computer-implemented method includes processing a query image by identifying one or more visual features in the query image and applying at least one nearest neighbor algorithm to the one or more identified visual features; identifying, from one or more databases, multiple images based at least in part on the processing; generating a result set by applying one or more smoothing algorithms to the multiple identified images; generating at least one sketch based at least in part on the result set; and outputting the at least one generated sketch to one or more users via a user interface.

Abstract: A neural network based search system is provided. A first digital image is analyzed by a user device. A targeted object in the first digital image is determined based, at least in part, on (i) the characteristics of the first digital image and (ii) the features of the targeted object. A vector array is generated based, at least in part, on (i) the first digital image and (ii) the targeted object. The vector array is analyzed by the user device. The targeted object is determined based, at least in part, by the vector array. A plurality of digital images is identified based, at least in part, on the similarity of the plurality of digital images and (i) the first digital image and (ii) the targeted object responsive to identifying a plurality of digital images, a query processing is generated. The query map is generated on a user device.