Microsoft Technology Licensing, LLC

Denis X Charles

Kristin E Lauter

Patrice Y Simard

David Maxwell Chickering

Leon Bottou

Kumar Hemachandra Chellapilla

Anton Mityagin

Carlos Garcia Jurado Suarez

David Max Chickering

David G Grangier

Saleema Amin Amershi

Aparna Lakshmiratan

Eyal Zvi Goren

Kamal Jain

Siddhartha Puri

Sven Seuken

Steven Mark Drucker

Bongshin Lee

Danyel Aharon Fisher

Sergey M Yekhanin

Reid Andersen

Sivakanth Gopi

Brandon J Maslen

Erik Daniel Anderson

Reinier M Broker

Rich Caruana

Joseph J Pfeiffer, Jr

Davis Leland Gilton

Todd Kulesza

Patrice Yvon Rene Simard

Aleksandr Rebrikov

John Robert Mooring

Disclosed is a system that tracks website usage without compromising user privacy. The system aggregates website usage data of multiple users across multiple websites. Website usage data is aggregated in a way that preserves each individual user's privacy. Specifically, information relevant to a particular user may be obtained from the aggregated information without exposing what was actually collected from that user. In some configurations, user-specific website usage data is aggregated using trusted execution environment computing hardware. This ensures that privacy is preserved while user-specific data is transferred to and processed by the system. The trusted execution environment applies differential privacy techniques to ensure that use of the aggregated information does not reveal actual information about a user's website usage history. In this way, privacy is maintained while still enabling many of the scenarios that would otherwise rely on third-party cookies.

Privacy preserving user personalization using noisy ranking

A system, method, and media are provided for generating a structured labeling graphical user interface. The user interface receives user input that associates multimedia content with categories. The user input may include user-supplied tags that further define the category for the multimedia content. The user-supplied tags are rendered proximate to the categories. In turn, a database logs user events to store, among other things, the categories, the user-supplied tags, time associated with completing the user-supplied tags, and time for associating multimedia content with the categories or tags.

Structured labeling to facilitate concept evolution in machine learning

Methods, computer systems, computer-storage media, and graphical user interfaces are provided for visualizing a performance of a machine-learned model. An interactive graphical user interface includes an item representation display area that displays a plurality of item representations corresponding to a plurality of items processed by the machine-learned model. The plurality of item representations are arranged according to scores assigned to the plurality of items by the machine-learned model. Further, each of the plurality of item representations is visually configured to represent a label assigned to a corresponding item.

Interactive visualization of machine-learning performance

A collection of data that is extremely large can be difficult to search and/or analyze. Relevance may be dramatically improved by automatically classifying queries and web pages in useful categories, and using these classification scores as relevance features. A thorough approach may require building a large number of classifiers, corresponding to the various types of information, activities, and products. Creation of classifiers and schematizers is provided on large data sets. Exercising the classifiers and schematizers on hundreds of millions of items may expose value that is inherent to the data by adding usable meta-data. Some aspects include active labeling exploration, automatic regularization and cold start, scaling with the number of items and the number of classifiers, active featuring, and segmentation and schematization.

Feature completion in computer-human interactive learning

Active labeling for computer-human interactive learning

Interactive segment extraction in computer-human interactive learning

Various embodiments provide techniques for graph clustering. In one or more embodiments, a participation graph is obtained that represents relationships between entities. An auxiliary graph is constructed based on the participation graph. The auxiliary graph may be constructed such that the auxiliary graph is less dense than the participation graph and is therefore computationally less complex to analyze. Clusters in the auxiliary graph are determined by solving an objective function defined for the auxiliary graph. Clusters determined for the auxiliary graph may then be utilized to ascertain clusters in the participation graph that solve a related objective function defined for the participation graph.

Graph clustering

Providing a market design for a peer-to-peer resource exchange system. Prices for a plurality of resources such as storage space, upload bandwidth, and download bandwidth are calculated and balanced based on previous resource prices, a supply of the resources, and a demand for the resources. Further, prices for operations such as storage and retrieval are determined such that a total of the payments to resource suppliers equals a total of the payments received from the resource consumers. In some embodiments, incoming data operation requests are allocated to the peers such that equilibrium among the peers is achieved.

Market design for a resource exchange system

Counterfactual analysis can be performed 'offline', or 'after the fact', based on data collected during a trial in which random variations are applied to the output of the system whose parameters are to be the subject of the counterfactual analysis. A weighting factor can be derived and applied to data collected during the trial to emphasize that data obtained when the random variations most closely resembled the output that would be expected if counterfactual parameters were utilized to generate the output. If the counterfactual parameters being considered differ too much from the parameters under which the trial was conducted, the offline counterfactual analysis can estimate a direction and magnitude of the change of the system performance, as opposed to deriving a specific expected system performance value. In economic transactions, the random variations can be considered variations in the price paid by another party, thereby enabling derivation of their marginal cost.

Offline counterfactual analysis

Systems and methods for computing modular polynomials modulo large primes are described. In one aspect, the systems and methods generate l-isogenous elliptic curves. A modular polynomial modulo a large prime p is then computed as a function of l-isogenous elliptic curves modulo p. In one aspect, the modular polynomial may be used in a cryptosystem.

Computing modular polynomials modulo large primes

Techniques are disclosed for representing and evaluating large prime degree isogenies for use in cryptographic signature and encryption schemes. An isogeny of prime degree 1 may be represented as an ideal in the form (1, A*alpha+B), where 1 comprises the degree of a prime number, the prime number is split into integers a and b, and alpha is a known endomorphism. For a given degree 1, integers a and b define a unique isogeny, allowing the isogeny to be stored with 3 log(1) bits of information. Techniques are also disclosed to evaluate the isogeny at a given point by decomposing the isogeny into an integer and a plurality of smaller degree isogenies, evaluating the smaller degree isogenies at the point with traditional means, and multiplying the results of the evaluations together and with the integer.

Cryptographic applications of efficiently evaluating large degree isogenies

An exemplary method includes receiving a request to register a peer in a peer-to-peer system; generating or selecting a transaction key for the peer; storing the transaction key in association with registration information for the peer; transmitting the transaction key to the peer and, in response to a request to perform a desired peer-to-peer transaction by another peer, generating a token, based at least in part on the transaction key. Such a token allows for secure transactions in a peer-to-peer system including remote storage of data and retrieval of remotely stored data. Other exemplary techniques are also disclosed including exemplary modules for a peer-to-peer server and peers in a peer-to-peer system.

Security architecture for peer-to-peer storage system

Systems and methods are described for trapdoor pairing. In one implementation, a trapdoor pairing is a cryptographic primitive generated by determining a bilinear pairing between an elliptic curve group and another group and selecting a parameter of the bilinear pairing, such as a group order or an isogeny between curves, to be a key for generating and evaluating the bilinear pairing. Trapdoor pairing allows construction of a group in which the Decisional Diffie-Hellman (DDH) problem is computationally infeasible given only the description of the group, but is easy given the secret key. Exemplary trapdoor pairing constructions have general applicability to cryptography and also lend themselves more specifically to certain special practical implementations, such as public key cryptography and certificate authority infrastructures.

Trapdoor pairings

Growing community of inventors

Bellevue, WA, United States of America

Denis X Charles