Abstract: A method of providing results from a search engine comprises generating an initial set based on a query received from a user; generating an attention value based on the query and on the initial set; applying the attention value to the initial set; and presenting a set of results in response to the query, wherein the results are generated according to the application of the attention value to the initial set.

Abstract: A method for machine learning-based classification may include training a machine learning model with a full training data set, the full training data set comprising a plurality of data points, to generate a first model state of the machine learning model, generating respective embeddings for the data points in the full training data set with the first model state of the machine learning model, applying a clustering algorithm to the respective embeddings to generate one or more clusters of the embeddings, identifying outlier embeddings from the one or more clusters of the embeddings, generating a reduced training data set comprising the full training data set less the data points associated with the outlier embeddings, training the machine learning model with the reduced training data set to a second model state, and applying the second model state to one or more data sets to classify the one or more data sets.

Abstract: A customizable router manages traffic between application programming interfaces (APIs) of microservice applications (apps) that make up various features of a website. Various datastores may also be used to store information, such as information related to certain products, pricing, users, etc. In order for the website to function properly, the microservice apps communicate with one another and with the datastores. The customizable router is used to manage and route traffic between the microservice apps and/or datastores. The customizable router may, for example, route a universal resource indicator (URI) request for a webpage of a website to particular versions or revisions of a microservice app based on rules established for the customizable router. For example, a certain percentage of traffic may be routed to a first version of a microservice app, while a different percentage of traffic may be routed to a second version of the microservice app.

Abstract: A method for generating a related product interface portion in an electronic user interface based on user-generated text includes identifying, for each product of a plurality of products, one or more respective characteristics from end user-generated text associated with the product, whereby a plurality of characteristics are identified for the plurality of products. The method further includes determining a plurality of characteristic groups, each group comprising two or more of the plurality of characteristics, wherein the characteristics within a group are similar to each other; receiving, from a user through the electronic user interface, an input related to a product category; and in response to receiving the user input, generating and presenting the related product interface portion that includes a respective product from each of the two or more of the characteristic groups related to the category.

Abstract: A computer-implemented method for determining image similarity includes determining, by a first neural network, a first feature value associated with a first characteristic of a first product based on an image of the first product. The method also includes determining, by a second neural network, a second feature value associated with a second characteristic of the first product based on the image of the first product. The method further involves calculating a first vector space distance between the first feature value and a third feature value associated with the first characteristic of a second product, and calculating a second vector space distance between the second feature value and a fourth feature value associated with the second characteristic of the second product. Additionally, the method includes determining a similarity value based on the first vector space distance and the second vector space distance.

Abstract: Systems and methods for distributing discrete events into a stateless cloud provider architecture are disclosed. Certain embodiments involve receiving, by the load balancer, a scaling up indication for the IVR system, the scaling up indication comprising an identification of a new server to be added to the plurality of servers. The method involves responsive to receiving the scaling up indication, registering the new server for IVR traffic by the load balancer. The method includes after registering the new server, receiving an IVR event. The method includes distributing the IVR event to the new server.

Abstract: Named Entity Recognition (NER) in a user search query in a real-time search engine may be achieved by training a machine learning algorithm to create a trained model. The trained model may be configured to receive a user search query as input and to recognize and output zero or more named entities in the search query. The training may include an iterative training process in which further training data is added at each iteration, in some embodiments. The training may be based on three training data sets, in some embodiments. A first training data set may be based on user search and engagement activity. A second training data set may be artificially generated based on a catalog of named entity values. A third training data set may be based on optimized and supplemented data pairs sampled from the first training data set.

Abstract: A method for returning a results page responsive to a user search query, such as a search query on a web site, may include receiving a search query from a user, determining, responsive to the query, a set of relevant products from a plurality of product listings based on a similarity of the user query to the respective product listings, retrieving inventory information respective of each of the relevant products, the inventory information comprising one or more available fulfillment channels respective of each of the relevant products, ranking the relevant products with respect to each other according to the inventory information, and returning, to the user, a search result comprising a list of the relevant products, ordered according to the ranking.

Abstract: A method includes determining a first taxonomy of an anchor product. The first taxonomy includes a plurality of levels for classifying products organized from a highest taxonomy level to a lowest taxonomy level. The method further includes determining a second taxonomy closest to the first taxonomy. The second taxonomy is associated with a group of products, the first taxonomy and the second taxonomy have at least a common highest taxonomy level, and the determination is made at least in part based on co-purchase data indicating that the anchor product and at least one product in the group of products are purchased together more often than products associated with other taxonomies are purchased with the anchor product. The method further includes determining a most similar product to the anchor product from the group of products of the second taxonomy and associating the anchor product and the most similar product with one another in a product collection.

Abstract: A method for ranking documents in search results includes defining a first training data set, the first training data set including, for each of a plurality of user queries, information respective of a document selected by a user from results responsive to the query and information respective of one or more documents within an observation window after the selected document in the results, and defining a second training data set, the second training data set including, for each of the plurality of user queries, information respective of the selected document. The method further includes training a first machine learning model with the first training data set, training a second machine learning model with the second training data set, and ranking documents of a further search result set according to the output of the first machine learning model and the output of the second machine learning model.

Abstract: An air filtration ceiling fan apparatus includes a ceiling fan unit and an air filtration unit. The ceiling fan unit includes a motor and a plurality of fan blades operably coupled to the motor. The air filtration unit includes a filter element and a filter fan configured to move air through the filter element. The ceiling fan unit is operably connected to the air filtration unit to form the air filtration ceiling fan, and the air filtration ceiling fan is configured to attach to a ceiling.

Abstract: An embodiment of a support assembly for a grill may include a leg subassembly comprising a first leg and a second leg, a leg subassembly comprising a first leg and a second leg, the first leg defining a retaining aperture, and a rotatable pin collar assembly configured to couple the first leg to the second leg in a slidable and lockable connection, the rotatable pin collar assembly including a spring pin configured to be retained in the retaining aperture at a first longitudinal position on the first leg when the grill is in a standing locked configuration and configured to be positioned at a second longitudinal position on the first leg when the grill is in a collapsed configuration.

Abstract: A light clamp apparatus includes a clamp configured to affix the light clamp apparatus to an object. The light clamp apparatus also includes a light assembly configured to removably attach to the clamp at least at two different positions on the clamp, the light assembly including a magnet, whereby the light assembly is removably attachable to a surface comprising a ferromagnetic metal. The clamp includes a slide bar comprising a first end and a second end, a first jaw coupled to the first end, and a body slidably receiving the slide bar having a second jaw opposing the first jaw. The clamp further includes a drive assembly for moving the slide bar in a first direction.

Abstract: A method of operating a price estimation system includes obtaining service requests from consumers, and obtaining bids from service professionals based on the service requests obtained from the consumers. The method additionally includes generating a training set based on the bids and the service requests, and generating a model for generating price estimates based on the training set. The method also includes employing the model to generate the price estimates based on additional service requests provided by additional consumers. In some aspects, the method further includes communicating the price estimates to the additional consumers during performance of a process for obtaining the additional service requests from the additional consumers. In other aspects, the method further includes communicating the price estimates to additional service professionals during performance of a process for obtaining additional bids from the additional service professionals based on the additional service requests.

Abstract: Systems and methods for providing controlled access to a system by a user device include receiving, from a user device, a request including a current context. The method includes receiving a request for access to a computing resource, the request including a current context, the current context defining a user space and a resource space. The user device evaluates the current context against a security policy. The user device determines that the user device is permitted to access the computing resource based on the request in response to the evaluating the current context against the security policy. In response to determining that the user device is permitted to access the computing resource, accessing the computing resource as requested.

Abstract: A method for operating a search engine may include determining a multi-dimensional search parameter space comprising a set of possible weight values for each of a plurality of search parameters and dividing the search parameter space into a grid of evenly-spaced values that is a subset of the set of possible values. The method may further include defining one or more initial populations of search parameter weight values, wherein each population of search parameter weight values comprises a plurality of initial individuals, wherein each initial individual comprises a respective one of the evenly-spaced values for each of the search parameters. The method may further include executing one or more genetic algorithms based on the one or more initial populations to select a final set of search parameter weight values, and returning results of a user search in the search engine according to the final set of search parameter weight values.

Abstract: A method includes receiving a plurality of product query arrays each including a plurality of individual product queries received during a single user search session. The method further includes inputting the plurality of product query arrays into the query rewrite model. Text of each of the plurality of individual product queries in each product query array is treated as a whole token. The method further includes receiving a product query from a user electronic device. The method further includes determining a query rewrite for the product query using the query rewrite model and determining search results for the product query using the query rewrite. The method further includes sending information for presenting the search results on a display of the user electronic device responsive to the product query.

Abstract: A method for machine learning-based classification may include training a machine learning model with a full training data set, the full training data set comprising a plurality of data points, to generate a first model state of the machine learning model, generating respective embeddings for the data points in the full training data set with the first model state of the machine learning model, applying a clustering algorithm to the respective embeddings to generate one or more clusters of the embeddings, identifying outlier embeddings from the one or more clusters of the embeddings, generating a reduced training data set comprising the full training data set less the data points associated with the outlier embeddings, training the machine learning model with the reduced training data set to a second model state, and applying the second model state to one or more data sets to classify the one or more data sets.