Abstract: Systems and methods are directed to automatic listing generation and inventory management based on machine-learning analysis. The system trains a time series-based machine learning (ML) model that forecasts sales. During inference time, the system determines one or more potential categories for a user based on custom preferences and previous analytic queries of the user. High demand items in the one or more potential categories are then applied to the ML model, which outputs probabilities of predicted sales for the high demand items. The system then determines items having a potential inventory gap by cross-checking current inventory with items having a probability outputted by the ML model that satisfies a probability threshold.

Abstract: Visual search query intent extraction and search refinement is described. In one or more implementations, a visual search query system receives a search query for items listed on an online marketplace, the search query including an image. Using one or more machine learning models, the visual search query system analyzes the image to determine characteristics of an object in the image. Based on the characteristics of the object in the image, the visual search query system automatically generates one or more search terms and searches the online marketplace to locate items matching the one or more search terms. The visual search query system then displays visual indications of the located items matching the one or more search terms in a user interface of the online marketplace.

Abstract: A multimodal embedding modifier generates a modified seed search selection embedding for providing a set of search results. The multimodal embedding modifier enhances the ability and accuracy of identifying a user's true intent when searching the online marketplace. For example, embodiments disclosed herein can allow a user to navigate multiple modalities for an item. In some embodiments, a user may select a search result corresponding to an initial search query, and further modify the selected search result by inputting a modifier (e.g., a textual modifier). The multimodal embedding modifier can be trained using a training dataset including a text embedding, an image embedding, another type of embedding, or a combination thereof.

Abstract: A multimodal embedding modifier generates a modified seed search selection embedding for providing a set of search results. The multimodal embedding modifier enhances the ability and accuracy of identifying a user's true intent when searching the online marketplace. For example, embodiments disclosed herein can allow a user to navigate multiple modalities for an item. In some embodiments, a user may select a search result corresponding to an initial search query, and further modify the selected search result by inputting a modifier (e.g., a textual modifier). The multimodal embedding modifier can be trained using a training dataset including a text embedding, an image embedding, another type of embedding, or a combination thereof.

Abstract: Systems and methods are directed to automatic listing generation and inventory management based on machine-learning analysis. The system trains a time series-based machine learning (ML) model that forecasts sales. During inference time, the system determines one or more potential categories for a user based on custom preferences and previous analytic queries of the user. High demand items in the one or more potential categories are then applied to the ML model, which outputs probabilities of predicted sales for the high demand items. The system then determines items having a potential inventory gap by cross-checking current inventory with items having a probability outputted by the ML model that satisfies a probability threshold.

Abstract: A multimodal embedding modifier generates a modified seed search selection embedding for providing a set of search results. The multimodal embedding modifier enhances the ability and accuracy of identifying a user's true intent when searching the online marketplace. For example, embodiments disclosed herein can allow a user to navigate multiple modalities for an item. In some embodiments, a user may select a search result corresponding to an initial search query, and further modify the selected search result by inputting a modifier (e.g., a textual modifier). The multimodal embedding modifier can be trained using a training dataset including a text embedding, an image embedding, another type of embedding, or a combination thereof.