Abstract: Automatically detecting and anticipating that an additional machine learning experiment may be needed. A method includes after successfully running a first experiment workflow, automatically prompting a user that an additional experiment workflow may be needed based on specific criteria associated with the first experiment workflow. The method further includes receiving input from the user confirming the additional experiment workflow. As a result of receiving input from the user confirming the additional experiment workflow, the method further includes the system automatically reconfiguring the first experiment workflow, including automatically identifying all necessary modules for the additional experiment workflow and connecting them properly to perform the intended second experiment workflow. The method further includes displaying to the user the first experimental workflow transitioning from the first experiment workflow to the additional experiment workflow.

Abstract: Organizations may comprise a variety of resources that are assigned to various organizational units, and allocated for various projects or tasks. A user who wishes to understand the organizational structure of the resources may examine information sources, such as resource manifests and resource access logs, and interpret such information with the user's knowledge of the organization, but may fail to achieve a comprehensive, accurate, and up-to-date understanding of the organization. Instead, interactions among users and resources may be monitored to infer organizational structure; assignment of resources to organizational units; relationships among resources; and usages arising within such resource relationships. Based on such inferences, an interactive visual map may be automatically generated that depicts the arrangement of the organization and the interrelationships of the resources.

Abstract: Mismatches between schema elements of a data set and a job are identified automatically. Furthermore, the mismatches can be presented visually in conjunction with an interactive visual workspace configured to support diagrammatic authoring of data transformation pipelines. After a data set is connected to a job, one or more mismatches can be determined and presented in context with the workspace. In addition, schema elements can be reconfigured by way of interaction with a visual representation of schema elements to resolve mismatches.

Abstract: Job execution can be scheduled and monitored. Execution of a job results in generation of one or more data slices, which can be consumed as input by one or more related jobs. Data slices can be presented in an interactive view. Selection of a data slice in the view can trigger automatic identification of related data slices based on dependencies between data slices. Each data slice can also include and be presented with a status indicating the availability of corresponding data, such as pending, ready, or failed. Furthermore, selection of a data slice that failed to generate successfully can trigger identification of related data slices that failed or are predicted to fail to generate successfully.

Abstract: Automatically detecting and anticipating that an additional machine learning experiment may be needed. A method includes after successfully running a first experiment workflow, automatically prompting a user that an additional experiment workflow may be needed based on specific criteria associated with the first experiment workflow. The method further includes receiving input from the user confirming the additional experiment workflow. As a result of receiving input from the user confirming the additional experiment workflow, the method further includes the system automatically reconfiguring the first experiment workflow, including automatically identifying all necessary modules for the additional experiment workflow and connecting them properly to perform the intended second experiment workflow. The method further includes displaying to the user the first experimental workflow transitioning from the first experiment workflow to the additional experiment workflow.

Abstract: Automatically detecting and anticipating that an additional machine learning experiment may be needed. A method includes after successfully running a first experiment workflow, automatically prompting a user that an additional experiment workflow may be needed based on specific criteria associated with the first experiment workflow. The method further includes receiving input from the user confirming the additional experiment workflow. As a result of receiving input from the user confirming the additional experiment workflow, the method further includes the system automatically reconfiguring the first experiment workflow, including automatically identifying all necessary modules for the additional experiment workflow and connecting them properly to perform the intended second experiment workflow. The method further includes displaying to the user the first experimental workflow transitioning from the first experiment workflow to the additional experiment workflow.

Abstract: Job execution can be scheduled and monitored. Execution of a job results in generation of one or more data slices, which can be consumed as input by one or more related jobs. Data slices can be presented in an interactive view. Selection of a data slice in the view can trigger automatic identification of related data slices based on dependencies between data slices. Each data slice can also include and be presented with a status indicating the availability of corresponding data, such as pending, ready, or failed. Furthermore, selection of a data slice that failed to generate successfully can trigger identification of related data slices that failed or are predicted to fail to generate successfully.

Abstract: A view of data transformation jobs can be presented by way of a user interface. Related jobs can subsequently be identified automatically after a job is selected based on data dependencies between jobs. Execution status can also be determined and presented such that successful and failed execution of jobs, for example, can be differentiated. Furthermore, selection of a job run that failed to execute successfully can trigger identification of related jobs runs that failed or are predicted to fail to execute successfully.

Abstract: Organizations may comprise a variety of resources that are assigned to various organizational units, and allocated for various projects or tasks. A user who wishes to understand the organizational structure of the resources may examine information sources, such as resource manifests and resource access logs, and interpret such information with the user's knowledge of the organization, but may fail to achieve a comprehensive, accurate, and up-to-date understanding of the organization. Instead, interactions among users and resources may be monitored to infer organizational structure; assignment of resources to organizational units; relationships among resources; and usages arising within such resource relationships. Based on such inferences, an interactive visual map may be automatically generated that depicts the arrangement of the organization and the interrelationships of the resources.

Abstract: While interacting with resources such as databases and data sets, a user of an organization may create a usage of a resource that is interesting or has value. The user may publish the usage to a web page and/or notify other users who may be interested in the usage, but the effectiveness of such sharing depends upon the user's volition to share the usage with interested users. Instead, the interaction between the interacting user and the resource may be monitored to detect a usage of the resource that may be interesting to other users of the organization. Interested users may be automatically identified and notified of the resource usage, and may request a replay of the resource usage by the interacting interested user. Usages may also be automatically stored as part of a catalog of resource usages by the interacting user for interested users to browse and replay.

Abstract: The resources of an organization may provide various functions and services, and the performance of the resources may be of interest. Performance indicators (e.g., network bandwidth, I/O throughput, and processing capacity) may be measured and logged, and examined in the event of diminished performance in order to identify a non-performing resource. However, because resources may depend on other resources, and to different degrees, the performance of the first resource may be limited by the performance of a related resource, and logging respective resource performance may not inform this determination. An embodiment of the techniques presented herein may evaluate an effect of the performance indicators of a first resource on the performance indicators of related resources (e.g., through historic comparisons or differential testing), and a visual resource map may include a visual indicator of the effect of performance indicators of a first resource on the performance indicators of a related resource.

Abstract: A view of data transformation jobs can be presented by way of a user interface. Related jobs can subsequently be identified automatically after a job is selected based on data dependencies between jobs. Execution status can also be determined and presented such that successful and failed execution of jobs, for example, can be differentiated. Furthermore, selection of a job run that failed to execute successfully can trigger identification of related jobs runs that failed or are predicted to fail to execute successfully.

Abstract: Automatically detecting and anticipating that an additional machine learning experiment may be needed. A method includes after successfully running a first experiment workflow, automatically prompting a user that an additional experiment workflow may be needed based on specific criteria associated with the first experiment workflow. The method further includes receiving input from the user confirming the additional experiment workflow. As a result of receiving input from the user confirming the additional experiment workflow, the method further includes the system automatically reconfiguring the first experiment workflow, including automatically identifying all necessary modules for the additional experiment workflow and connecting them properly to perform the intended second experiment workflow. The method further includes displaying to the user the first experimental workflow transitioning from the first experiment workflow to the additional experiment workflow.

Abstract: Mismatches between schema elements of a data set and a job are identified automatically. Furthermore, the mismatches can be presented visually in conjunction with an interactive visual workspace configured to support diagrammatic authoring of data transformation pipelines. After a data set is connected to a job, one or more mismatches can be determined and presented in context with the workspace. In addition, schema elements can be reconfigured by way of interaction with a visual representation of schema elements to resolve mismatches.

Abstract: Jobs can be created within a visual authoring environment. A new job of a selected type can be added to a diagrammatic workspace. Subsequently, a mechanism configured to enable selection of a saved job that implements all or a portion of the job can be presented. After selection, a saved job can be acquired and the workspace updated based thereon. Furthermore, data sources associated with the saved job can be can be added to a data source designated portion of the environment.

Abstract: Jobs can be authored in conjunction with a visual workspace. Upon selection of representation of a data source in the workspace, a preview of the data source can be generated within context of the visual workspace. Further, representations of one or more data transformation operations can be provided with the preview. Selection of a transformation operation results in an updated preview reflecting application of the operation as well as generation of backend code to perform the operation. Furthermore, a job comprising one or more transformation operations can be added to the workspace automatically.

Abstract: Methods, systems and architectures for enhancing a user's DVD experience are described. In one embodiment, the physical structure (e.g. titles and chapters) of a DVD is ascertained and a playlist structure shell that represents the DVD's physical structure is built. Metadata associated with the DVD is acquired and used to populate the playlist structure shell. In some embodiments, the metadata can be acquired from a metadata source external to the DVD and/or remote from a computing device on which the DVD is being played. Remote locations can include, without limitation, those that are network- or Internet-accessible. Once the playlist structure shell is populated, it can be used to render a playlist associated with the DVD's video in an area of a user interface display.

Abstract: Methods, systems and architectures for enhancing a user's DVD experience are described. In one embodiment, the physical structure (e.g. titles and chapters) of a DVD is ascertained and a playlist structure shell that represents the DVD's physical structure is built. Metadata associated with the DVD is acquired and used to populate the playlist structure shell. In some embodiments, the metadata can be acquired from a metadata source external to the DVD and/or remote from a computing device on which the DVD is being played. Remote locations can include, without limitation, those that are network- or Internet-accessible. Once the playlist structure shell is populated, it can be used to render a playlist associated with the DVD's video in an area of a user interface display.

Abstract: Methods, systems and architectures for enhancing a user's DVD experience are described. In one embodiment, the hierarchical structure (e.g. titles and chapters) of a DVD is ascertained and a corresponding playlist structure shell is built. Metadata associated with the DVD is acquired and used to populate the playlist structure shell. In some embodiments, the metadata can be acquired from a metadata source external to the DVD and/or remote from a computing device on which the DYD is being played. Remote locations can include, without limitation, those that are network- or Internet-accessible. Once the playlist structure shell is populated, it can be used to render a playlist associated with the DVD's video in an area of a user interface display.

Abstract: The present invention provides a system and method for improving a user's ability to direct the computer to perform a desired function. A single uniform input control is provided that receives user input, such as from a keyboard. The present embodiment then parses the user input to determine its input type and whether the input type corresponds to a single, recognized action performable by the computer and, if so, performs the action. Such recognized actions include navigating to an Internet site and composing an e-mail message. If the input type does not correspond to a single recognized action, the user input is submitted to at least one text processor, such as a natural language processor, that analyzes the user input to obtain at least one interpretation corresponding to an action performable by the computer. The at least one interpretation is returned to the user who may select a desired action from among the listing of interpretations.