Abstract: Methods and systems for preventing unauthorized resource access. In some aspects, the system obtains, in real time, a first data stream between a user requesting access to resources in a user account and an agent. The system processes, using a first machine learning model, the first data stream to generate a confidence score. If the confidence score exceeds a threshold, the system generates a sandbox context configured to simulate output indicative of successful resource access. The system obtains a second data stream for the sandbox context and processes it using a second machine learning model to determine whether the communication between the agent and the user was malicious. If the communication was not malicious, the system removes the user account from the sandbox context and effects grant of resource access. Otherwise, the system reports the agent and the user account for further processing for attempts to obtain unauthorized user access.

Abstract: Systems and methods for detecting anomalous data updates in federated learning. In some aspects, the system retrieves a plurality of data updates and a corresponding plurality of data profiles during a first interval of time. The system generates a first data profile trend based on the plurality of data profiles, including an expectation value and a measure of variance. The system receives a subsequent data update and a corresponding subsequent data profile. The system determines a measure of deviation based on the expectation value of the first data profile trend and the subsequent data profile. The system generates an anomaly score based on the measure of deviation and the measure of variance. Based on the anomaly score, the system determines whether to label the subsequent data update as acceptable for inclusion in training data to generate a local model.

Abstract: Methods, systems, and apparatuses are described herein for an extension that predicts a user's Visually Impaired Spectrum (VIS) score then adjusts a browser accessibility setting accordingly. Further, the method may monitor the user's interaction with the accessibility settings and update the user's predicted VIS score. An extension may train a machine learning model to predict a user's VIS score, then generate a VIS score employing the user's personal information as input. Further, the extension may retrieve, from a database, one or more accessibility settings associated with a VIS score. The extension may adjust one or more browser accessibility settings according to the VIS score.

Abstract: Disclosed herein are system, method, and computer program product embodiments for generating a set of correlated applications by a recommendation system to resolve an incident for an application service (AS). The recommendation system can determine an AS-to-AS similarity matrix based on incident data and contextual data, an AS-to-AS affinity matrix based on the incident data and contextual data, and generate a set of correlated applications based on the similarity matrix and the affinity matrix. The set of correlated applications can also be generated based on pairwise application associations, where an application association between a first application and a second application indicates that the first application and the second application can occur together in an incident or to be changed together as indicated by a change record.

Abstract: A method, a system, and a computer program product for using contactless cards and mobile applications for selecting and using methods of payment for transactions. An identifier associated with a transaction is received. The transaction is configured to be executed using at least one computing device. One or more virtual identifiers are generated using the received identifier. At least one virtual identifier is selected and assigned to the transaction. Execution of the transaction is triggered by the computing device using the assigned virtual identifier.

Abstract: Methods, systems, and apparatuses are described herein for associating past auction bidding activity with subsequent financial transactions. A computing device may receive, from a web browser plugin of a web browser application executing on a user computing device, bid information relating to bids placed by a user on an auction website. Such information may be collected by, for example, scraping the Document Object Model of the website. After the auction ends, the computing device may determine a plurality of different candidate financial transactions for the user. Based on an estimated payment amount, the computing device may select, from the plurality of different candidate financial transactions, a matching financial transaction, then output an association between a bid and the matching financial transaction.

Abstract: Systems and methods for adaptive configuration. In some aspects, a system may receive an indication that a communication session is initiated between a public terminal and a user device in proximity of the public terminal. In response to receiving the indication, the system may obtain historical transaction data of the user. Based on the historical transaction data, the system may determine (1) frequent locations of the user where a user has conducted more than a threshold number of transactions and (2) a language preference for the user. In response to determining that the user is not located within one or more frequent locations of the user, the system may transmit a command to alter a digital display of the public terminal based on a user-specific configuration, wherein the user-specific configuration comprises the language preference of the user.

Abstract: Systems and methods for detecting anomalous data updates in federated learning. In some aspects, the system receives a plurality of data updates. Each data update contains data in a first real-valued space. The system selects a first function to project the plurality of data updates into a second real-valued space. The system selects a second function to partition the second real-valued space into a plurality of sectors. The system generates a plurality of sector datasets associated with the plurality of sectors. The system processes the plurality of sector datasets to generate a relational data structure. The system determines outliers in the relational data structure corresponding to anomalous data updates.

Abstract: Disclosed herein are system, method, and computer program product embodiments for verifying a document. An embodiment operates by receiving an initial image of a document and identifying a first imperfection on a feature of the document. Thereafter, the feature of the initial image is segmented into a first and second partially overlapping segment corresponding to a first and second portion of the feature, respectively. Subsequently, the feature's first portion is determined to be free from the first imperfection, and the first imperfection is identified in the feature's second segment. After receiving a second image of the document including the feature's second portion, the feature's second portion in the second image is determined to be free from a second imperfection. As a result, an image of the feature is created based on the first and second portions of the feature derived from the initial and second images, respectively.

Abstract: Methods and systems for generating synthetic data. In some aspects, the system receives a first set of candidate behaviors, each of which comprises plain text. The system processes the first set of candidate behaviors using a first language processing model to generate a set of representations in an embedding space. For each representation in the set of representations, the system processes the representation using a second language processing model to generate a first sequence of behavioral tokens representative of a timeline of user activities. Using the set of representations and first sequence of behavior tokens, the system updates the second language processing model. Using the updated second language processing model, the system processes a second set of candidate behaviors to generate a second sequence of behavior tokens.

Abstract: Systems and methods for detecting abusive behavior. In some aspects, a method includes receiving first transfer data for a first transfer associated with a second category of resources and a first indication whether the first transfer is associated with abusive behavior. The first transfer data may be processed using a first machine learning model to generate a first prediction regarding whether the first transfer is associated with abusive behavior. In response to determining that the first prediction does not match the first indication, the first machine learning model may be updated to generate a second machine learning model for the second category of resources. The method may include receiving second transfer data for a second transfer associated with a second category of resources and processing the second transfer data using the second machine learning model to generate a second prediction regarding whether the second transfer is associated with abusive behavior.

Abstract: Systems and methods for providing a separate interest rate for an individual transaction are disclosed. A system may provide a service associated with a account having an account balance and an associated first interest rate. The system may include a service device having one or more memory devices storing instructions and one or more processors configured to execute the instructions to perform a separate interest rate service process. The process may include providing a separate interest rate service offer to a client device and receiving a separate interest rate selection from the client device. The process may also include determining a selected portion of the account balance corresponding to an amount to which a second interest rate, different from the first interest rate, is to be applied based on the separate interest rate selection.

Abstract: Multidimensional data analysis applications, including OLAP applications, simultaneously aggregate across many sets of dimensions. However, computing multidimensional aggregates is a performance bottleneck for OLAP data analysis applications. In order to improve the speed of interactive analysis, OLAP databases often precompute aggregates at various levels of detail and on various combinations of data attributes. However, the cost and speed of precomputation influences how frequently the aggregates can be brought up-to-date. Systems and methods disclosed herein provide graph based multidimensional analysis processing without pre-aggregating or precomputing the data along dimensional hierarchies, and by providing the results to the end user on-demand. Since preaggregation or precomputation of data along dimensional hierarchies is not necessary, implementations allow the end user to perform data analysis as soon as the data is available.

Abstract: In some implementations, a device of a user may receive user input indicative of a selection of an application implemented by the device. The user input may be received via a graphical user interface presented by the device. The user input may be received while the application is in an unlaunched state in which the user is unauthenticated for use of the application. The device may obtain, while the application is in the unlaunched state, one or more messages of the application. The device may cause, while the application is in the unlaunched state, information indicative of at least one message, of the one or more messages, to be displayed in the graphical user interface.

Abstract: Disclosed embodiments accelerate web form completion based on email address analytics. An email address can be input with respect to a web form and identified. The email address can be parsed into a username and a domain name and analyzed. A value of a web form field can be determined based on data extracted from the username, domain name, or both. The determined value can then be employed to automatically complete a corresponding web form field Values for additional fields can be computed based on the determined value alone or combined with data from other sources. User input and interaction with a web form can be monitored and used as a basis to dynamically complete additional fields, alter fields, or clear field values.

Abstract: An automated teller machine (ATM) has a display and a wireless communication interface. The ATM may detect a proximity of a mobile computing device and exhibit on its display an indication of the proximity of mobile computing device to the ATM. The ATM may also be configured to detect when the mobile computing device and the ATM have established a wireless communication connection and display a visual cue on the display that the wireless connection has been established. The ATM is configured to synchronize the display of the ATM with a display of the mobile computing device. The displays may be synched temporally and positionally and may display the same or related content, such as an image or animation at the same time.

Abstract: Disclosed embodiments may include a system that may receive, from a user device associated with a user, authentication rule(s), each authentication rule associated with respective merchant(s) and comprising respective required authentication factor(s) for completing a transaction with the respective merchant(s). The system may identify a user has navigated to a merchant webpage and may determine at least one rule of the authentication rule(s) associated with the merchant. The system may identify first required authentication factor(s) corresponding to the at least one rule, and may cause the user device to display, via a GUI, the first required authentication factor(s). The system may receive, via the user device, authentication information associated with the user. The system may determine whether the authentication information satisfies the first required authentication factor(s). Responsive to making that determination, the system may enable the user to complete a transaction via the merchant webpage.

Abstract: Aspects described herein may allow an automated generation of an interactive multimedia content with annotations showing vehicle damage. In one method, a server may receive vehicle-specific identifying information of a vehicle. Image sensors may capture multimedia content showing aspects associated with exterior regions of the vehicle, and may send the multimedia content to the server. For each of the exterior regions of the vehicle, the server may determine, using a trained classification model, instances of damage. Furthermore, the server may generate an interactive multimedia content that shows images with annotations indicating instances of damage. The interactive multimedia content may be displayed via a user interface.

Abstract: In some implementations, a monitoring device may receive configuration information associated with an extract, transform, load (ETL) pipeline that includes one or more data sources and one or more data sinks. The monitoring device may generate, based on the configuration information, lineage data related to a data flow from the one or more data sources to the one or more data sinks in the ETL pipeline. The monitoring device may generate one or more predicted quality metrics associated with the ETL pipeline using a machine learning model. The monitoring device may generate a visualization in which multiple nodes are arranged to indicate the data flow from the one or more data sources to the one or more data sinks and further in which the one or more predicted quality metrics are encoded within the visualization.