公开(公告)号：US10250587B2

公开(公告)日：2019-04-02

申请号：US15282656

申请日：2016-09-30

Applicant: Microsoft Technology Licensing, LLC.

Inventor： Elad Iwanir ,  Gal Tamir ,  Shahar Weiss ,  Eli Koreh

IPC: H04L29/06

Abstract: The misuse of public key, private key, and public/private key certificates poses significant security challenges to computer networks that are addressed by certificate monitoring. Certificate monitoring allows network administrators to detect and remedy poor security practices related to public key certificates and to detect and combat the malicious use of public key certificates in a centralized environment. Best practices and detection methods and systems are developed over time via machine learning to improve network security, and any detected misuse may be brought to a network administrator's attention or automatically remedied.

公开(公告)号：US20180097803A1

公开(公告)日：2018-04-05

申请号：US15282656

申请日：2016-09-30

Applicant: Microsoft Technology Licensing, LLC.

Inventor： Elad Iwanir ,  Gal Tamir ,  Shahar Weiss ,  Eli Koreh

IPC: H04L29/06

CPC classification number: H04L63/0823 ,  H04L63/0815 ,  H04L63/14

Abstract: The misuse of public key, private key, and public/private key certificates poses significant security challenges to computer networks that are addressed by certificate monitoring. Certificate monitoring allows network administrators to detect and remedy poor security practices related to public key certificates and to detect and combat the malicious use of public key certificates in a centralized environment. Best practices and detection methods and systems are developed over time via machine learning to improve network security, and any detected misuse may be brought to a network administrator's attention or automatically remedied.

公开(公告)号：US10558458B2

公开(公告)日：2020-02-11

申请号：US15174688

申请日：2016-06-06

Applicant: Microsoft Technology Licensing, LLC

Inventor： Elad Iwanir ,  Gal Tamir ,  Amir Eluk ,  Eli Koreh

IPC: G06F8/72 ,  G06F16/2453

Abstract: Methods, systems, apparatuses, and computer program products are provided for increasing an efficiency of queries in program code. A plurality of queries is detected in program code. A laziness is extended by which the queries are evaluated in the program code. The queries are decomposed into a plurality of query components. A ruleset that includes a plurality of rules is applied to the query components to generate a functionally equivalent query set to the plurality of queries that evaluates more efficiently relative to the plurality of queries.

公开(公告)号：US20170351512A1

公开(公告)日：2017-12-07

申请号：US15174688

申请日：2016-06-06

Applicant: Microsoft Technology Licensing, LLC

Inventor： Elad Iwanir ,  Gal Tamir ,  Amir Eluk ,  Eli Koreh

IPC: G06F9/44 ,  G06F17/30

Abstract: Methods, systems, apparatuses, and computer program products are provided for increasing an efficiency of queries in program code. A plurality of queries is detected in program code. A laziness is extended by which the queries are evaluated in the program code. The queries are decomposed into a plurality of query components. A ruleset that includes a plurality of rules is applied to the query components to generate a functionally equivalent query set to the plurality of queries that evaluates more efficiently relative to the plurality of queries.

公开(公告)号：US10943181B2

公开(公告)日：2021-03-09

申请号：US14751135

申请日：2015-06-26

Applicant: Microsoft Technology Licensing, LLC

Inventor： Hanan Shteingart ,  Yair Tor ,  Eli Koreh ,  Amit Hilbuch ,  Yifat Schacter

IPC: G06N20/00

Abstract: Disclosed herein is a system and method that can be used with any underlying classification technique. The method receives a test dataset and determines the features in that test dataset that are present. From these features the training dataset is modified to only have those features that are present in the test dataset. This modified test dataset is then used to calibrate the classifier for the particular incoming data set. The process repeats itself for each different incoming dataset providing a just in time calibration of the classifier.

公开(公告)号：US20160379135A1

公开(公告)日：2016-12-29

申请号：US14751135

申请日：2015-06-26

Applicant: Microsoft Technology Licensing, LLC

Inventor： Hanan Shteingart ,  Yair Tor ,  Eli Koreh ,  Amit Hilbuch ,  Yifat Schacter

IPC: G06N99/00

CPC classification number: G06N20/00

Abstract: Disclosed herein is a system and method that can be used with any underlying classification technique. The method receives a test dataset and determines the features in that test dataset that are present. From these features the training dataset is modified to only have those features that are present in the test dataset. This modified test dataset is then used to calibrate the classifier for the particular incoming data set. The process repeats itself for each different incoming dataset providing a just in time calibration of the classifier.

Abstract translation: 本文公开的是可以与任何基础分类技术一起使用的系统和方法。 该方法接收测试数据集并确定存在的测试数据集中的特征。 从这些功能中，训练数据集被修改为仅具有测试数据集中存在的那些特征。 然后，修改后的测试数据集用于校准特定输入数据集的分类器。 该过程对于每个不同的传入数据集重复，从而提供分类器的正确的时间校准。

公开(公告)号：US10157055B2

公开(公告)日：2018-12-18

申请号：US15280754

申请日：2016-09-29

Applicant: Microsoft Technology Licensing, LLC

Inventor： Gal Tamir ,  Elad Iwanir ,  Eli Koreh

IPC: G06F9/45 ,  G06F8/72 ,  G06F8/41

Abstract: Methods, systems, apparatuses, and computer program products are provided for transforming asynchronous code into more efficient, logically equivalent asynchronous code; Program code is converted into a first syntax tree. A dependency graph is generated from the first syntax tree with each node of the dependency graph corresponding to a code statement and having an assigned weight. Weighted topological sorting of the dependency graph is performed to generate a sorted dependency graph. A second syntax tree is generated from the sorted dependency graph. In another implementation, the program code is transformed into await-relaxed and/or loop-relaxed program code prior to being transformed into the first syntax tree.

公开(公告)号：US20180088937A1

公开(公告)日：2018-03-29

申请号：US15280754

申请日：2016-09-29

Applicant: Microsoft Technology Licensing, LLC

Inventor： Gal Tamir ,  Elad Iwanir ,  Eli Koreh

IPC: G06F9/44 ,  G06F9/45

CPC classification number: G06F8/72 ,  G06F8/433 ,  G06F8/4441 ,  G06F8/458

Abstract: Methods, systems, apparatuses, and computer program products are provided for transforming asynchronous code into more efficient, logically equivalent asynchronous code; Program code is converted into a first syntax tree. A dependency graph is generated from the first syntax tree with each node of the dependency graph corresponding to a code statement and having an assigned weight. Weighted topological sorting of the dependency graph is performed to generate a sorted dependency graph. A second syntax tree is generated from the sorted dependency graph. In another implementation, the program code is transformed into await-relaxed and/or loop-relaxed program code prior to being transformed into the first syntax tree.

公开(公告)号：US20160379133A1

公开(公告)日：2016-12-29

申请号：US14748211

申请日：2015-06-23

Applicant: Microsoft Technology Licensing, LLC

Inventor： Hanan Shteingart ,  Yair Tor ,  Eli Koreh ,  Amit Hilbuch ,  Yifat Schacter

IPC: G06N99/00

CPC classification number: G06N20/00

Abstract: Disclosed herein is a system and method that can be used with any underlying classification technique. The method takes into account both the value of the current feature vector. It is based on evaluating the effect of perturbing each feature by bootstrapping it with the negative samples and measuring the change in the classifier output. To assess the importance of a given feature value in the classified feature vector, a random negatively labeled instance is taken out of the training set and replaces the feature at question with a corresponding feature from this set. Then, by classifying the modified feature vector and comparing its predicted label and classifier output a user is able measure and observe the effect of changing each feature.

Abstract translation: 本文公开的是可以与任何基础分类技术一起使用的系统和方法。 该方法考虑了当前特征向量的值。 它是基于通过用负样本引导来扰动每个特征的效果，并测量分类器输出的变化。 为了评估给定特征值在分类特征向量中的重要性，从训练集中取出随机负面标记的实例，并用该集合中的相应特征替换所讨论的特征。 然后，通过对修改的特征向量进行分类并比较其预测标签和分类器输出，用户可以测量和观察改变每个特征的效果。