Microsoft Technology Licensing, LLC

Muiris Woulfe

Andreas Balzer

David Mowatt

Alan Noel Mulhall

Stephen O'Driscoll

Poornima Muthukumar

Yuanyuan Dong

Johnny S Campbell

Rolando Jimenez Salgado

Venkat Pradeep Chilakamarri

The description relates to an interactive physical product browsing experience. One example includes a display system for displaying objects. The display system comprises a plurality of object displays each comprising an object store, a robot for moving objects to and from a repository, a user input receiving means, and a processing means. The processing means is configured to cause the object displays to display some of the objects, monitor user behaviour using the user input receiving means, and cause the robot to move an object from the repository to one of the object stores based on the user behaviour.

Interactive physical product browsing experience

Techniques for rendering shared data include receiving, from a data store by a computing device, data indicative of a persistent data object. The persistent data object is associated with a class indicative of a data type for information contained in the persistent data object. The persistent data object is operable to be inserted in a file generated by an application executing on the computing device. The visualization logic associated with the class may be received by the computing device. The information in the persistent data object is rendered using the selected method of visualization, and other information in the document is rendered using native rendering capabilities of the application.

Visualization platform for reusable data chunks

A machine learning model can be trained to infer the probability of the presence of categories of a software bug in a source code file. A bug tracker can provide information concerning the category to which a software bug belongs. The bug data supplied to a machine learning model for inferring the presence of particular categories of bugs can be filtered to exclude a specified category or categories of bugs. Information including but not limited to organizational boundaries can be inferred from the category of bugs present in a body of source code. The inferred organization boundaries can be used to generate team-specific machine learning models.

Bug categorization and team boundary inference via automated bug detection

Methods and systems which perform information retrieval using natural language dialogue for navigating an inventory of items are described. One example provides an information retrieval system to a user using natural language dialogue. The system comprises a user input receiving device, an output device, a database comprising an inventory of items, and a processor. The processor is configured to retrieve one or more items from the inventory of items using an iterative process by: in response to receiving from the user input receiving device a user input, identifying a subset of the inventory based on the user input. The processor is configured to automatically process the subset of items to determine a classification for distinguishing between items of the subset, to generate an enquiry for a user using the classification and to transmit the enquiry to the output device. The user input and/or the enquiry may use natural language.

Information retrieval using natural language dialogue

A deep learning model is trained on historical pull requests to automatically identify appropriate reviewers to review source code from one or more source code repositories. The model is trained on features that are based on past pull requests from the source code repositories and that represent the context of the syntactic representation of the changed code. The model learns patterns found in the changed source code and of the past peers associated with the changed source code to relate certain source code fragments with certain peers. The model generates probabilities based on the learned patterns which are used to identify appropriate reviewers more suitable to review the source code.

Automatic identification of appropriate code reviewers using machine learning

Techniques for managing data include receiving, at a data store, a persistent data object generated by a source application, the object configured to be compatible with a plurality of applications and document types. The object is associated with a unique identifier. In response to a request for the object, the object is accessed based on its unique identifier and sent to a computing device executing a destination application. The object is incorporated by and is compatible with a destination document being edited by the destination application. An update to the object is received that is generated by a user application editing a user document. In response to receiving an indication that the object has been inserted in the destination document, the update is sent by the data store to the destination computing device and is usable to update the object as incorporated in the destination document.

Data re-use across documents

An unmanned aerial vehicle (UAV) rendezvous with and transfers a product to a receiving vehicle that is en route to a destination-location. The UAV is dispatched with the product along a flight path that intercepts with a predetermined route that the receiving vehicle is expected to travel along toward the destination-location. Once the UAV is within the vicinity of the receiving vehicle, the UAV approaches the receiving vehicle and utilizes cargo release equipment to transfer the product to the receiving vehicle. In one example, the UAV flies above the receiving vehicle at a synchronized velocity and drops the product through an opening in the roof of the receiving vehicle. In another example, the UAV flies above the receiving vehicle and suspends the product adjacent to a side-window opening of the receiving vehicle to enable an occupant of the receiving vehicle to reach out and retrieve the product.

En route product delivery by unmanned aerial vehicles

A system includes an electronic processor configured to store records in a client database. The records included personally identifiable information associated with entities and a client identifier reference associated with each of the individuals, the personally identifiable information is accessible based on an authorization level associated with a user. The electronic processor is also configured to receive a document containing both non-personally identifiable information and personally identifiable information, parse the document to determine the presence personally identifiable information in the document, verify the access rights of the reader of the document, match the personally identifiable information included in the document with records stored in the client database, substitute the personally identifiable information included in the document with the corresponding client identifier reference referring to the client database, and generate a human-legible view of the document containing the corresponding client identifier reference.

Protecting personally identifiable information (PII) using tagging and persistence of PII

The description relates to determining tiles of interest to a user. One example includes a computing device comprising a display, a user input receiving means, and a processing means. The processing means is configured to render on the display a first plurality of tiles each having a respective value of a first classification. The processing means is configured to receive from the user input receiving means a first user input indicating user interest in a first tile of the first plurality of tiles, the first tile having a first value of the first classification. The processing means is configured, in response to receiving the first user input, to render on the display a second plurality of tiles each having a respective value of a second classification and being related to the first tile by having a value of the first classification within a threshold similarity of the first value.

Adaptive tile-based user interface for inferring user interest

A system for protecting personally identifiable information (PII) associated with audio, image and video. The system includes an output device and a processor. The processor receives a document including an audio, an image, or a video containing both non-personally identifiable information and personally identifiable information, scans the document for a voice, a face, a graphically rendered text, or a personal attribute, match the voice, face, graphically rendered text, or personal attribute with records in a database to determine whether the voice, face, graphically rendered text, or personal attribute in the document is associated with personally identifiable information. The processor also determines a start time and an end time associated with the presence of the voice or video in response to determining that the voice, or video is associated with PII, generates an obfuscated audio or a video between the start time and the end time, and causes the output device to output the obfuscated audio, graphically rendered text or video.

Obfuscating information related to personally identifiable information (PII)

An increasing number of bots become available each day that perform automated tasks over the Internet to help facilitate a variety of service requests for a user. Thus, embodiments are directed to an orchestration service configured to perform bot network orchestration in order to provide enriched responses to service requests and/or in order to find one accurate answer among large numbers of bot responses to a specific query. For example, a service request for a user that involves at least one service provider may be received. A plurality of bots to orchestrate a processing of the service request may be determined. Instructions may be provided to the bots associated with aspects of the service request, where each bot is selected based on an aspect of the service request. Responses received from the bots may be integrated into a service response and provided to a requestor of the service request.

Bot network orchestration to provide enriched service request responses

Information concerning software bugs including bug detection, bug prediction data and/or historical bug data can be used to determine whether it is safe to commit, integrate, deploy and/or deliver a software change. If the change is deemed unsafe, the change can be rejected automatically. Alternatively, the change can proceed following approval by an administrator, supervisor, implementer, manager and/or other designated approval mechanism. Actions taken to override a block can be recorded along with information concerning failed deployments and/or deliveries, a higher than normal customer failure rate after deployment and/or delivery or through manual data entry.

Engineering system robustness using bug data

For each detected bug, historical code with similar characteristics and bug corrections from a historical bug dataset can be displayed in a source code editor. Relevant training and/or testing data can be found by comparing an internal representation of the code under development with an internal representation of the original buggy code in the historical bug dataset. Training and/or testing data that is relevant to the current code can be distinguished from irrelevant training and/or testing data by determining that the code syntax tokens from the current and historical data overlap to at least a specified percentage. Code can be devolved into a set of metrics. The degree of overlap between the metric sets can be determined. If a computed risk factor for the bug correction meets or exceeds a specified threshold, the bug correction can be automatically applied. Additional testing can be automatically performed on and/or added to the corrected code.

Enhancing software development using bug data

Technologies are provided for automated crime scene analysis using machine learning. Firearm models, types, or even specific firearms may be automatically detected from captured audio files or continuous audio streams (e.g., recording microphones) using machine learning techniques. The detection may also be based on (or enhanced by) captured still images or video files/streams. Further information such as crime scene layout, wound types and locations, and similar information may be provided to the analysis service through manual input or automated capture (e.g., through analysis of image/video data). A number of firearms used in the commission of the crime may also be detected. Specific firearm types may be associated with specific crime types. Similar techniques may also be used to detect and classify types and quantity of explosive material.

Crime scene analysis using machine learning

Methods and systems which perform information retrieval using natural language dialog for navigating an inventory of items are described. One example provides an information retrieval system to a user using natural language dialog. The system comprises a user input receiving device, an output device, a database comprising an inventory of items, and a processor. The processor is configured to retrieve one or more items from the inventory of items using an iterative process by: in response to receiving from the user input receiving device a user input, identifying a subset of the inventory based on the user input. The processor is configured to automatically process the subset of items to determine a classification for distinguishing between items of the subset, to generate an enquiry for a user using the classification and to transmit the enquiry to the output device. The user input and/or the enquiry may use natural language.

Growing community of inventors

Dublin, Ireland

Muiris Woulfe