Abstract: A machine learning (“ML”) pipeline that includes unsupervised learning, supervised learning, and Bayesian learning is utilized to train a ML classifier that can classify machine metrics as being indicative of an anomaly. A boosting process can be utilized during the unsupervised learning portion of the ML pipeline that scores clusters of training data for completeness, and further splits clusters of training data based upon the completeness scores in order to optimize the clustering of the training data. Supervised learning is then performed on the cluster-labeled training data. Bayesian learning can also be utilized to assign incident probability inferences to the clusters of training data. Once the ML classifier has been trained, the ML classifier can be utilized in a production environment to classify multi-dimensional machine metrics generated by computing devices in the production environment as being indicative of an anomaly.

Abstract: In various example embodiments, a system and method for providing dynamic bidding notifications and price guidance for buyers are presented. The system receives a selection of a present item listing and accesses historical auction data for a set of historical item listings. The system generates a dynamic bidding model for bidding on the present item and generates one or more bidding notifications based on the dynamic bidding model. The system then causes presentation of the one or more bidding notifications on a client device.

Abstract: Technologies are disclosed herein for enhancing machine learning (“ML”)-based anomaly detection systems using knowledge graphs. The disclosed technologies generate a connected graph that defines a topology of infrastructure components along with associated alarms generated by a ML component. The ML component generates the alarms by applying ML techniques to real-time data metrics generated by the infrastructure components. Scores are computed for the infrastructure components based upon the connected graph. A root cause of an anomaly affecting infrastructure components can then be identified based upon the scores, and remedial action can be taken to address the root cause of the anomaly. A user interface is also provided for visualizing aspects of the connected graph.

Abstract: A system, computer-readable storage medium storing at least one program, and computer-implemented method for automatically posting content to a network accessible content collection are described. User activity over a communication network from a client device is tracked and a triggering event is detected. In response to detecting the triggering event, a content item is automatically posted to a network accessible collection of content items viewable by additional users.

Abstract: In various example embodiments, a system and method for a Listing Engine that translates a first listing from a first language to a second language. The first listing includes an image(s) of a first item. The Listing Engine provides as input to an encoded neural network model a portion(s) of a translated first listing and a portions(s) of a second listing in the second language. The second listing includes an image(s) of a second item. The Listing Engine receives from the encoded neural network model a first feature vector for the translated first listing and a second feature vector for the second listing. The first and the second feature vectors both include at least one type of image signature feature and at least one type of listing text-based feature.

Abstract: The disclosed technologies include receiving a request from a second computing device to verify ownership of a blockchain address. A challenge content is generated and sent to the requestor. A signature is received comprising a hash of the challenge content generated using a private key. A public key corresponding to the private key is obtained, and the signature is validated using the public key. In response to validating the signature, a characteristic is associated with a user associated with the blockchain address.

Abstract: Techniques are disclosed for automatically adjusting machine learning parameters in an e-commerce system. Hyperparameters of a machine learning component are tuned using a gradient estimator and a first training set representative of an e-commerce context. The machine learning component is trained using the tuned hyperparameters and the first training set. The hyperparameters are automatically re-tuned using the gradient estimator and a second training set representative of a changed e-commerce context. The machine learning component is re-trained using the re-tuned hyperparameters and the second training set.

Abstract: In various example embodiments, heavy document object model (DOM) elements in a heavy web page are removed directly from a DOM tree subsequent to a request being received to navigate to another web page but before the web browser actually navigates to that other web page. In one example embodiment, the heavy DOM elements are deleted from the previous page during an ‘onbeforeunload’ event using a scripting language such as Javascript, to ensure that the render of the next page starts faster.

Abstract: In accordance with one or more embodiments of the present disclosure, a method for leveraging a social network to search for local inventory includes communicating with a first mobile communication device over a network, the first mobile communication device associated with a user, and communicating with a second mobile communication device over the network, the second mobile communication device associated with a social network contact. The method further includes obtaining a geo-location of the second mobile communication device, processing a search request from the first mobile communication device, obtaining a radius of coverage for a merchant, determining whether the geo-location of the second mobile communication device is within the radius of coverage of the merchant, and providing a listing to the first mobile communication device of one or more items for sale by the merchant when the second mobile communication device is within the radius of coverage of the merchant.

Abstract: Systems, computer-implemented methods, and computer-readable storage mediums for determining tour locations are disclosed. Users are provided access to items of an online marketplace of a first user. Data is received from the first user that includes a request to determine a tour location of a tour. User accesses to the items of the online marketplace are monitored. Based at least on the monitored user accesses, values for candidate locations are determined. One of the candidate locations are selected based at least on the determined values of the candidate locations. The selected location of the candidate locations may be provided to the first user.

Abstract: Methods and systems for displaying, to a user interface, categories of items for the user to select from for querying the system, in order to enable the system to return to the user images of items of the type desired by the user for possible purchase. The items categories of items can be presented to the user interface visually in silhouette form so that the user can select brand and the silhouette image of the type of item desired. Upon selection of the silhouette image, a query is sent to the system and images of the desired type of item are returned to the user interface for presentation to the user, along with certain attributes of the items.

Abstract: A web browser extension identifies graphic objects from images or video being presented by a web browser. Webpages related to the graphic objects are identified. Web links that facilitate navigation to the webpages are embedded over an area of the image corresponding to the identified graphic image. Where the graphic objects are identified within video, the web links are progressively embedded within graphic object boundaries of the graphic object as the graphic objects move locations during progression of the video. In this way, a user is able to interact with graphic objects of images and video to navigate to webpages related to the graphic objects. Some implementations provide a webpage redirect command at a stop point of the video so that the user can interact with graphic objects while the video is playing and without interrupting the video.

Abstract: In various example embodiments, a system and method for providing price guidance for sellers and buyers are presented. The system receives a present item listing and accesses a set of historical item listings. The system generates a price guidance model for the present item listing and generates a set of prices for the present item based on the price guidance model. The system then causes presentation of the set of process on a client device.

Abstract: Technologies are shown for secure management of evaluation data that involves receiving an evaluation value signal from a source, the evaluation value signal relating to an evaluation entity having an evaluation score secured on an evaluation data blockchain and verifying whether the source is identified in trusted source data. If the source is trusted, then the technology involves obtaining a weight associated with the source, obtaining the evaluation score for the evaluation entity from a first evaluation data block in the evaluation data blockchain, where the first evaluation data block is a most recent evaluation data block in the evaluation data blockchain, calculating a new evaluation score based on the evaluation score obtained from the first evaluation data block and the received valuation signal weighted according to the weight associated with the source, and securely committing the new evaluation score to the evaluation data blockchain in another evaluation data block.

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for listing conversion estimation and optimization. A listing conversion estimation and optimization system receives a set of listing data for generating a listing of an item for sale on an online marketplace and determines, based on the set of listing data, a likelihood value indicating the likelihood that the listing will result in a sale within a given time period. The listing conversion estimation and optimization system causes presentation of a graphical indicator representing the likelihood value on the display of the listing user's client device.

Abstract: In various example embodiments, a system and method for facilitating display of virtual content are presented. A session that displays two-dimensional (2D) content of one or more items available for sale is presented on a first device of a user. A second device of the user is detected, the second device being able to display three-dimensional (3D) content of the one or more items available for sale. 3D content of the one or more items available for sale is retrieved. Display of the 3D content on the second device is caused, the 3D content selectable by the user to perform interactions with the 3D content. An indication of the user performed interactions is received and processed. A result that depicts the user performed interactions as being processed is displayed on the first device of the user.

Abstract: An inventory prediction system is described that outputs a predicted inventory item not included in a user's known inventory using a cross-category directional graph that represents item categories as nodes. The inventory prediction system implements a prediction model trained using machine learning to output the predicted inventory item using the graph and at least one item from the user's known inventory. The inventory prediction system is further configured to generate a listing recommendation for the predicted inventory item. To do so, the inventory prediction system implements a logistic regression model trained using machine learning to calculate a probability that the listing recommendation should be generated using attributes of the predicted inventory item and attributes of currently trending items.

Abstract: A machine is configured to generate in real time personalized online banner images for users based on data pertaining to user behavior in relation to an image of a product. For example, the machine receives a user selection indicating one or more data features associated with the user. The one or more data features include a data feature pertaining to user behavior in relation to an image of a product. The machine generates, using a machine learning algorithm, a data representation of the machine learning algorithm based on the one or more data features including the data feature pertaining to user behavior in relation to the image of the product. The data representation includes one or more data features pertaining to one or more characteristics of online banner images. The machine generates an online banner image for the user based on the data representation.

Abstract: A counterfeit item detection system detects counterfeit items during an item listing processes provided by an online marketplace. The system enhances the ability of the online marketplace to identify and reject potential counterfeit items. The system collects item data in various formats. The item data is analyzed using speech-to-text software and natural language processing to determine data elements representing items, item features, and language context. Questions are generated using the items and item features, and stored for each item. Answers to the questions have associated counterfeit indication weights. The weights are modified and used to rank the questions as feedback is received about counterfeit items. The ranking determines future question selection, allowing the best questions to continually be identified and provided. The counterfeit item detection system also trains a neural network to detect counterfeit items in images, where the training images are obtained from videos related to the item.

Abstract: Methods, systems, and apparatus for monitoring a sound are described. An audio signal is obtained and the audio signal is analyzed to generate an audio signature. An object type is identified based on the audio signature and an action corresponding to the object type is identified.

Abstract: A system may receive information representing a first item and an item price of the first item from a first user device operated by a buyer. The first item may be available for sale by a merchant of a brick-and-mortar store. A plurality of sellers may be identified, the plurality of sellers may list a second item as being available for sale at the calculated reduced item price in an item listing for the second item. Moreover, the second item may be similar to the first item. A reduced item price may be calculated based on the item price of the first item received from the user device, the reduced item price representing a price lower than the item price of the first item. The item listing for the second item available for sale may be presented to the first user device operated by the buyer.

Abstract: Aspects of the present disclosure involve a system comprising a computer-readable storage medium storing at least one program, and a method for rendering interactive visualizations of high volume streaming data. Streaming data that describes a plurality of events is loaded in memory. The streaming data includes geospatial data that specifies a source location and a destination location for each event. Visual metaphors representing the events are rendered using a visualization mathematical model that defines, for each visual metaphor, a spatial position and opacity for each point along the visual metaphor throughout a sliding time window.

Abstract: Systems and methods of cloud deployment optimization are disclosed. In some example embodiments, a method comprises running original instances of an application concurrently on original servers to implement an online service, receiving, by the original instances of the application original requests for one or more functions of the online service, receiving a command to deploy a number of additional instances of the application, transmitting synthetic requests for the function(s) of the online service to one of the original servers according to a predetermined optimization criteria, deploying the number of additional instances of the application on additional servers using a copy of the original instance of the application, and running the deployed additional instances of the application on their corresponding additional servers concurrently with the original instances of the application being run on their corresponding original servers.

Abstract: Technologies are shown for application risk assessment in an authentication service where an authorization request is received from a third party application calling an Application Programming Interface (API). Risk assessment policies that pertain to behavioral characteristics, such as API usage patterns or past delegation of permissions, are applied to the authorization request to obtain a risk assessment score. If the risk assessment score does not exceed a risk threshold, then an authorization message is sent in response to the authorization request. If the risk assessment score exceeds the risk threshold, then remedial action, such as suspending the application, limiting the available actions, or sending a notification to a trusted security application, is executed for an account associated with the third party application. Machine learning can be applied to historical behavioral data to generate the risk assessment policies.

Abstract: Methods, systems, and apparatus for receiving one or more power usage measurements for a set of elements, determining a user behavior based at least on an analysis of the one or more power usage measurements of at least one of the set of elements, and based on a determined user behavior, causing the at least one of the set of elements to perform an action that affects a power function of the at least one of the set of elements.

Abstract: Systems and methods for multi-directional visual browsing on an electronic device are described. In example embodiments, a primary result and a peripheral result are determined. A display layout based on attributes associated with the primary result and the peripheral result is generated. The display layout is then formatted into instructions, which will cause a device to render the display layout. The instructions are transmitted to the client device.

Abstract: Different action user-interface components in a comparison view are described. Initially, a selection is received to display a comparison view via a user interface of a listing platform. Multiple listings of the listing platform are selected for inclusion in the comparison view. A comparison view system determines which action of a plurality of actions, used by the listing platform, to associate with each of the listings. A display device displays the multiple listings concurrently in a comparison view via a user interface of the listing platform and also displays an action user-interface component (e.g., a button) in each of the plurality of listings. The action user-interface component is selectable to initiate the action associated with the respective listing. In accordance with the described techniques, the action user-interface component displayed in at least two of the multiple listings is selectable to initiate different actions in relation to the respective listing.

Abstract: Real time item listing modification is described. A listing modification system receives, via a real time communication protocol from a client device, user interaction data that describes an item listing of a network-based commerce system currently being accessed at the client device. The user interaction data further describes an identity of a user of the client device and a mariner in which the user of the client device interacts with the item listing. A modification user interface is generated for output at a different computing device and configured to present the user interaction data together with controls that are selectable to cause modification of the item listing output to the client device. The modified item listing is selectively output to only the client device, such that other computing devices simultaneously accessing the item listing are not updated to display the modified item listing.

Abstract: In various example embodiments, a system and method for modifying publications of a network-based publication system are presented. The system receives a set of body coordinates from a client device and associates the set of body coordinates with a user account. The system identifies a set of clothing images within a set of publications retrieved as a result of a query. The system generates a set of modified clothing images from one or more images of the set of clothing images and causes presentation of the set of modified clothing images at the client device.

Abstract: Electronic content that has a tactile dimension when presented on a tactile-enabled computing device may be referred to as tactile-enabled content. A tactile-enabled device is a device that is capable of presenting tactile-enabled content in a manner that permits a user to experience tactile quality of electronic content. In one example embodiment, a system is provided for generating content that has a tactile dimension when presented on a tactile-enabled device.

Abstract: According to example embodiments, an Image View Aggregator identifies a frontal view of an item within an image. The Image View Aggregator identifies at least one reflection view of the item within the image. Each reflection view of the item having been captured off a corresponding reflective physical surface. The Image View Aggregator extracts the frontal view of the item and each reflection view of the item from the image. The Image View Aggregator generates a representation of the item based at least on the extracted frontal view of the item and each extracted reflection view of the item.

Abstract: A system comprising a computer-readable storage medium storing at least one program, and a computer-implemented method for providing digital avatars. A first request message is received for requesting a digital representation of a first target user combined with a target product for an online marketplace application on a first client device. The first request message is associated with a first user identifier indicative and a product identifier. Based on the first user identifier, a first user record is selected from a first database. The first user record includes measurement data. Based on the product identifier, a product record is selected from a second database. Based at least on the measurement data of the first user record and the selected product record, model data of the target product is determined and provided to the first client device.

Abstract: There are provided systems and methods for wireless beacon devices for use in tracking user locations during group travel. A user may take a trip on a vehicle, such as a bus, train, cruise ship, or other large travel vehicle. The users may disembark the vehicle in order to visit a location but plan to travel further on the vehicle. Thus, when the vehicle is attempting to leave the location where the users have left the vehicle, the vehicle may be required to again account for each passenger. The vehicle may utilize a wireless beacon located near an entrance and/or onboard the vehicle to account for each passenger. The wireless beacon may connect to a communication device of each user when the communication device is in proximity to the wireless beacon and, thus, inform the vehicle's operator that each user in located on the vehicle.

Abstract: A server accesses a plurality of users' sessions with the web server. Each user session indicating a page flow of a corresponding user session for a plurality of web pages provided by the web server. The server generates a learning model using a neural network based on the plurality of users' sessions. The learning model is configured to predict a next user activity based on a current page flow of a current user session. The next user activity indicating one of continuing the current user session by visiting another web page provided by the web server and ending the current user session. The server dynamically adjusts a content of a web page based on the predicted next user activity.

Abstract: Disclosed is a tamper proof tag and accompanying method for verifying that a watch is authentic and has not been tampered with. The tamper proof tag, when affixed to the watch, ensures that the watch is authentic and prevents a user from tampering with the watch. The tamper proof tag also allows the watch to be worn while the tamper proof tag is affixed to watch and is minimally obstructive to the aesthetics of the watch. This provides a purchasing user with confidence that the watch is authentic, while also allowing the user to try on the watch without having to remove the tamper proof tag. The tamper proof tag also protects the seller from a user that may attempt to tamper with the watch after purchase. For example, the seller may condition returns for a sold watch on the tamper proof tag remaining affixed to the watch.

Abstract: Technologies are shown for shipping route selection involving receiving sender and recipient shipping information for an item to be shipped and obtaining shipping route options for the item based on the sender and recipient shipping information. The route options are provided for display and selection of an option. A routing data block is created for the item at an address on a blockchain that stores shipping information for each stage of the selected route. A shipping tag is encoded with the blockchain address and attached to the item. The tag can be scanned to obtain the blockchain address and request information for a next stage of the shipping route from the block. The next shipping information from the block is received and utilized to ship the item to a next geolocation. The next shipping information can be determined based on current conditions, such as weather, pricing and availability.

Abstract: Listing an item on a publication system involves sending the publication system an email or other communication that includes title, description and a picture of the item. The email may be in plain text, with the publication system using the email for listing the item with minimal complexity for the user. The publication system may receive the email, parse the email to extract title, description, price, classification, attributes and perhaps other information with respect to the item, and compose a view item page (“VIP”) for the item listing. The extracting may also be by decoding a product identifier in the communication. The VIP may be privately viewable and editable by the user. The user may then modify or otherwise confirm the VIP. When the user confirms the VIP, the VIP goes live as the listing on the publication system.

Abstract: A system comprising a machine-readable storage medium storing at least one program and a computer-implemented method for collecting body measurements of a human user using a body measurement garment is provided. The body measurement garment comprises a plurality of sensors configured to produce a set of output data. The method may include receiving the set of output data from the body measurement garment and determining a plurality of body measurements from the output data. The method may further include generating a body shape model representing the shape of the body of the user. The method may also include generating a garment fit model representing the fit of a garment on the human user.

Abstract: Systems and methods for multi-signal fault analysis are described. The system receives signal message information, over a network, from a collection device comprised of a plurality of mobile devices including a first mobile device and a second mobile device that are associated with a first user. The first signal message information includes a first maintenance message including characterization information that was received by the first mobile device from a component that includes a sensor that operates to sense a first part that is assembled into a vehicle. The second signal message information includes signal information that was received by the second mobile device including an audio signal that a microphone in the second mobile device sensed. The system analyzes the multi-signal information to diagnose a problem and communicate a message to the first user with a diagnosis of the problem.

Abstract: Techniques image suitability determination to generate augmented reality/virtual reality (AR/VR) digital content are described. A two-dimensional digital image is received. Using machine-learning, a determination as to whether an object captured by the two-dimensional digital image is suitable for generating AR/VR digital content for display in an AR/VR environment is made. If the object is suitable, an indication is provided and an option to view the object in an AR/VR environment is provided. If the object is not suitable, a suggestion indicating why the object as captured is not suitable and/or how to correct the capture of the object in a subsequent digital image such that it is suitable for generating AR/VR digital content.

Abstract: In various example embodiments, a system and method for a proactive customer support system are provided. In some example embodiments, outgoing communications from an application server to a client device are monitored for error messages, outgoing error messages are detected, an error type for an error message is determined, an issue ticket including the error message and the error type is generated, and instructions are transmitted that cause a customer service device to display the issue ticket. In some example embodiments, the system additionally assigns a priority score and ranks open error tickets based on their respective error messages. In some example embodiments, the system provides a help message to the client device based on the error message, receives an information request from the client device, determines a reply message based on the information request, and transmits instructions to the client device to display the reply message.

Abstract: Colored three-dimensional digital model generation techniques and systems are described. In one example, scanning techniques are employed by a scanning system that scans a physical object while disposed within packaging to form a three-dimension digital model. A model coloring system is employed to color the three-dimensional digital model. A two-dimensional digital image is employed that captures the same or similar physical object. In one example, features of the model are matched to the image. This is then used to align a viewing perspective with respect to the model with a viewing perspective of the object within the digital image, e.g., to find which “view” of the model corresponds with the image. The color is then applied from the digital image to the model, e.g., from pixels of the image to corresponding points in the model.

Abstract: A resource provider is disclosed which, upon receiving selection of a resource from a first party for delivery to a third party, automatically determines configuration and/or customization settings for the specific resource based upon the third party user's target device and/or their preferred customizations. In doing so, the first party does not need to have specific knowledge about the third party in order to correctly configure or customize the resource for use by the third party. Upon receiving a selection of a resource, the resource provider determines configurable and customizable features for the selected resource.

Abstract: A method of enhancing a network-based publication system with curator lists is disclosed. Information pertaining to an item is received from a submitter of a listing of the item on a network-based publication system. Metadata pertaining to the information about the item is received from a curator of the information pertaining to the item. A presentation of the information on a curator list is controlled based on the metadata.

Abstract: Systems and methods are provided for providing access to secure-access facilities based on pairing of the secure-access facilities with a user device such as a wearable device. A pairable secure-access facility may be a public storage facility or device such as a locker that includes communications circuitry for pairing with the user device. Once paired with the user device, the locker may operate a locking mechanism to lock the locker when the user device is away from the locker and to unlock the locker when the user device is in the vicinity of the locker. The locker may include a beacon for detecting and pairing with the user devices. Pairing the user device and the locker may include entering a locker identifier into the user device to ensure that the intended user device is paired with the intended locker.

Abstract: Systems and methods for visualization of an item in an environment using augmented reality are provided. Environment image data containing an image of an environment is received. A selection of an item for placement into an indicated location of the environment is received. An item image of the selected item is scaled based on dimensions determined from the environment image data for the environment. The scaled item image is augmented into the image of the environment at the indicated location to generate an augmented reality image that also includes information for the item. The augmented reality image is displayed on a device of a user.

Abstract: In various example embodiments, a system and method for a cross selling platform are presented. A request for a product listing triggered by a user may be received. The request for the product listing may include product information corresponding to an in-store product located at a store of a merchant. A related product listing may be identified based on the product information. Each product listing of the plurality of product listings may correspond to a merchant product. The product listing may be recommended to the user for sale.

Abstract: A database object used in a plurality of database operations is determined. A live range of the database object is computed. The computing of the live range includes determining occurrences of the database operations to the database object. Based at least in part on the live range of the database object, a memory is determined to be optimally assigned to store the database object based on at least one characteristic of the memory. A first time to allocate the database object to the memory is determined. A second time to deallocate the database object from the memory is determined. An output file comprising a first instruction to store the database object in the memory at the first time and a second instruction to deallocate the database object from the memory at the second time is written.

Abstract: Systems, methods, and machine-readable hardware to facilitate a diagnosis of a failure of a first part in a vehicle. The system receives a maintenance message, over a network at a network-based marketplace. The maintenance message includes characterization information received from a component in the vehicle. The component includes a sensor that senses the first part to generate the characterization information. The system identifies sub-system processing rules based on the maintenance message and executes the sub-system processing rules. The sub-system processing rules cause a smart component engine to perform operations comprising identifying a sub-system and scanning the sub-system responsive to identifying an increase in a likelihood of failure of the first part based on the characterization information.

Abstract: Technologies are disclosed herein for cross-correlating metrics for anomaly root cause detection. Primary and secondary metrics associated with an anomaly are cross-correlated by first using the derivative of an interpolant of data points of the primary metric to identify a time window for analysis. Impact scores for the secondary metrics can be then be generated by computing the standard deviation of a derivative of data points of the secondary metrics during the identified time window. The impact scores can be utilized to collect data relating to the secondary metrics most likely to have caused the anomaly. Remedial action can then be taken based upon the collected data in order to address the root cause of the anomaly.