Microsoft Technology Licensing, LLC

Motorola Corporation

Cezary Antoni Marcjan

Lili Cheng

Andrzej Turski

Jerome D Boss

Matthew Bret MacLaurin

Fei Su

Ryszard K Kott

Gilad Odinak

David Milstein

Johnson T Apacible

Shoou-Jiun Wang

Richard F Rashid

Jayaram N M Nanduri

Shelly D Farnham

Sean U Kelly

Hung-Chih Yang

Mark A Nikiel

David P Vronay

David Shigeru Taniguchi

Yuting Jia

Surajit J Chaudhuri

Todd D Newman

Steven Wayne Ickman

Sean P Nolan

Michael W Thomas

David J Anderson

David Patrick Vronay

Jeffrey Dick Jones

Danny A Bersch

Tom Laird-McConnell

Howard C Anderson

Jayaram Nm Nanduri

Tien Widya Suwandy

Ming-Yu Fan

Huiying Mao

In non-limiting examples of the present disclosure, systems, methods and devices for training conversational language models are presented. An embedding library may be generated and maintained. Exemplary target inputs and associated intent types may be received. The target inputs may be encoded into contextual embeddings. The embeddings may be added to the embedding library. When a conversational entity receives a new natural language input, that new input may be encoded into a contextual embedding. The new embedding may be added to the embedding library. A similarity score model may be applied to the new embedding and one or more embeddings for the exemplary target inputs. Similarity scores may be calculated based on the application of the similarity score model. A response may be generated by the conversational entity for an intent type for which a similarity score exceeds a threshold value.

Natural language processing models for conversational computing

Training risk determination models based on a set of labeled data transactions. A first set of labeled data transactions that have been labeled during a review process is accessed. A first risk determination model is trained using the first set of labeled data transactions. A first risk score for data transactions of a set of unlabeled data transactions is determined using the first risk determination model. Data transactions in the set of unlabeled data transactions are newly labeled based on the first risk score. The newly labeled data transactions are added to a second set of labeled data transactions that include the first set of labeled data transactions. A second risk determination model is trained using at least the second set of labeled data transactions. A second risk score is determined for subsequently received data transactions and these data transactions are rejected or approved based on the second risk score.

Using semi-supervised label procreation to train a risk determination model

Embodiments disclosed herein are related to computing systems and methods for determining a risk score for a plurality of data transactions. In the embodiments, a first risk score module may receive data transactions. The first risk score module may then determine a first risk score for each of the data transactions. A second risk score module that is different from the first risk score module may receive each of the first risk scores determined by the first risk score module as an input. The second risk score module may determine a second risk score based in part on the input first risk scores for each of the data transactions. The second risk scores may specify if each of the data transactions is to be approved or rejected by the computing system.

Long-term short-term cascade modeling for fraud detection

Embodiments disclosed herein are related determining a risk score for one or more data transactions. Current data transactions that are associated with one or more current attributes are received. Stored data transactions associated with stored attributes are accessed. A plurality of the stored attributes are selected. A first sliding window and a second sliding window are selected. A duration of the second sliding window is longer than a duration of the first sliding window and encompasses the duration of first sliding window. Risk information for those stored data transactions that are associated with the plurality of attributes is determined. The risk information is determined during the duration of both the first and second sliding windows and is indicative of a level fraud that is occurring. The determined risk information and the current attributes are used to generate a risk score for the current data transactions. The current data transactions are approved or rejected based on the risk score.

Hierarchical profiling inputs and self-adaptive fraud detection system

A resource identifier parser can derive information from resource identifiers that are provided to non-browser application programs or the operating system, such as through the desktop area of a windows-based user interface. The resource identifier parser can understand the formatting of resource identifiers of specific domains and can derive therefrom information such as coordinates of a map page, the title of classified listings, and other like information. If the resource identifier cannot be parsed, the identified data page can be referenced to identify services offered by the page, such as data feed services, which can then be presented to the non-browser application program or the operating system to which the resource identifier was provided. As yet another alternative, the domain can provide a customized resource identifier parser to parse its resource identifiers. Such a customized parser can be either downloadable and locally executable or it can be a network service.

Application integration of network data based on resource identifiers

In one example, data may be divided into blocks, and the blocks may be stored in various storage resources. Data to be stored may be provided to a data divider. The data divider may divide the data into blocks. Redundancy may be introduced into the blocks to allow the original data to be reconstructed at some point in the future, even if fewer than all of the original blocks are available. The blocks may be sent to a data shuffler, which may shuffle the blocks, and may provide a key that describes how to reconstruct the original data from the blocks. The key may be provided to the owner of the data. When the original data is to be retrieved, the key may be provided to a reassembler, which retrieves the blocks from the various storage resources and reconstructs the data using the key.

Distributed data storage

A state affinitized system comprises worker computing devices uniquely affinitized with one or more states such that only those worker computing devices maintain local copies of the states with which they are affinitized. Requests and data relevant to a state are directed to the worker computing device affinitized with that state, such as by front-end computing devices. A centralized persistent storage can maintain a copy of each state and can modify the states in response to newly received information in the same manner as the worker computing devices. Such maintenance can be performed by coalescer computing devices, which can also derive second order data from the states in the centralized persistent storage, which can be further added to the state and also communicated to the worker computing device affinitized with that state. Crawler computing devices can obtain additional information for incorporation into states from network-based information repositories and services.

Scalable affinitized state management

A system described herein includes a receiver component that receives a computing event with respect to a file stored in a data repository of a personal computing device. An assignor component automatically assigns a uniform resource identifier to the file responsive to the receiver component receiving the computing event with respect to the file, wherein at least a portion of content of the file is reviewable by way of a web browser through utilization of the uniform resource identifier.

Sharing files on a computer through use of URIs

A database query user interface combines the user convenience of simple text searching with the expressive refinements of powerful query language. The database query user interface includes a query text string input from a user including one or more terms of a chunk expression language format. The database query user interface further includes a syntactical prompt for constructing a multi-element chunk expression language database query that is syntactically correct and complete and includes the text string input from the user. For example, the syntactical prompt is selected from the database based upon a weighted analysis of database information relating to database elements included in the text string input from the user. A database query formed according to the present user interface may then be persisted or stored as a database query object.

Database query user interface to assist in efficient and accurate query construction

A protocol is provided for communicating data between two applications. The protocol can support communication of an extensible markup language (XML) document over hypertext transfer protocol (HTTP) by associating the XML document with a data envelope and sending the data envelope to a disparate application. The data envelope can comprise a header, which can specify supported compression methods, encryption keys, and/or data specific to the disparate application and/or the sending application. The protocol adds a layer of security to the communication and can prevent malicious requests and other attacks. This protocol can also be used in conjunction with a health integration network.

Health integration platform protocol

Data may be stored in a location or manner that takes various considerations into account. Examples of such considerations are the availability, speed and cost of storage resources, and commercial and regulatory expectations concerning the reliability, security, and/or availability of the stored data. When a piece of data is to be stored, a storage choice component may take the above-considerations into account in order to determine where the data is to be stored, how many copies of the data are to be made, whether the data is to be encrypted, and/or other issues. Additionally, a migration engine may re-evaluate data that has already been stored in order to determine whether the data may be migrated to other storage resources, and/or whether changes may be made with regard to issues such as the encryption level and/or the number of stored copies of the data.

Choosing location or manner of storing data

Systems and/or methods that mitigate overexposure of shared content and that facilitate improving security of content in a computing environment are provided. In particular, the systems and methods involves monitoring resource activities such as user interactions (between users and/or between users and resources), email activity, share space activity, shared content activity, distribution lists, contact lists, personnel or organization structures, and the like. Data can be collected as one or more resource activities are monitored and then analyzed to determine where, if any, changes have taken place. For example, if a user has been removed from a distribution list or a user has not accessed the shared content within a time period, a recommendation can be formulated to suggest an adjustment to the affected access rights. Furthermore, the systems and/or methods can be integrated with other applications such as an email client, whereby an out-of-office indicator can leave access rights as they are.

Privacy model that grants access rights and provides security to shared content

A computer system models human memory by deriving associations between objects, events, and the context of the computer user or users. These associations can be dynamically generated, changing depending on the behavior of the user and context. Examples of areas in which this system can be used include time management (e.g., a calendar that presents time-based groupings of objects), people management (e.g., use analysis of user communications to rank importance of people to the user, groupings, and associations between people/groups and other objects such as documents, email), and general computer management (e.g., use analysis of user behavior to identify important objects and objects that are related to a current focus and context of the computer user).

Computer system architecture for automatic context associations

A search of an index database or another search method is conducted to identify preliminary results listing one or more selected computer objects having selected identifying information stored in an index database. In addition, one or more selected computer objects of the preliminary search results are correlated with one or more other computer objects that have associations with the selected computer objects of the preliminary search results. Integrated search results are then returned and include the preliminary search results and one or more other computer objects that have associations with the selected computer objects of the preliminary search results. The associations may be determined by a association system and represent relationships between computer files based upon user or other interactions between the objects. The associations between the objects may include similarities between them and their importance.

System and method for searching computer files and returning identified files and associated files

A unique system, user interface, and method that facilitate sharing objects in a query-based storage environment are provided. A query can be generated using metadata and the results therefrom can be obtained and shared with one or more recipients. The query and/or query results can be represented as an icon. Sharing of the query and/or query results can be accomplished in part by dragging the icon to the desired recipient. Queries can be shared as a standing query or a snapshot of the query. For standing queries, recipients can be notified when a set of query results changes. In general, queries can be saved by the user and accessed again at a later time. The user interface provides a plurality of navigation tools to facilitate the viewing of objects based on their metadata and the generation of queries.

Growing community of inventors

Redmond, WA, United States of America

Cezary Antoni Marcjan