Microsoft Technology Licensing, LLC

Ron Jack Karidi

Roy Varshavsky

Moshe Tennenholtz

Shai Rubin

Oded Elyada

Lior Arzi

Avigad Oron

Efim Hudis

Benny Schlesinger

Eyal Krupka

Eran Yariv

Hen Fitoussi

Yair Helman

Eugene (John) Neystadt

Arie Friedman

Daniel Sitton

Kira Radinsky

Gal Lavee

Limor Lahiani

Ziv Rafalovich

Adar Greenshpon

Liat Segal

Rotem Bennett

Ofer Dekel

Noga Amit

Yitzhak Tzahi Weisfeild

Systems, methods, and computer media for integrating requests for reserved allocations of advertisement impressions and requests for dynamic allocations of advertisement impressions are provided. A request is received from a first advertiser to purchase a reserved allocation of advertisement impressions. The request specifies a requested number of impressions that each have one or more requested attributes. One or more bids to dynamically purchase one or more impressions through a real-time bidding process are received from an external bidding agent. An internal bidding agent bids, on behalf of the first advertiser, to dynamically purchase one or more impressions through the real-time bidding process until the request from the first advertiser to purchase the reserved allocation of impressions is satisfied.

Integration of reserved and dynamic advertisement allocations

A ranked set of users may be calculated from an expertise categorization for each user and a person's trust in the users for specific categories. The ranked set of users may be used for presenting search results, recommendations, social marketing, or other uses. A person's reputation may be determined through various online activities. A person's trust in another person may be related to their proximity and activity in one or more social networks.

Social network based contextual ranking

Techniques for determining a set of keywords associated with a document are provided. A document is received that may be classified into a taxonomy that includes a plurality of categories. A categorization ranking is determined for each category for the received document. A set of categories of the taxonomy having highest categorization rankings is determined for the received document. Documents representing the set of categories having highest categorization rankings are combined together into a cumulative representative text that includes a plurality of terms. A cumulative term corpus importance score is determined for each term in the cumulative representative text. The cumulative term corpus importance score for a particular term indicates an importance of the particular term in a context of the cumulative representative text. A set of terms of the cumulative representative text having highest cumulative term corpus importance scores is selected to be keywords for the received document.

Keywords extraction and enrichment via categorization systems

Topical affinity badges in information retrieval is described, for example, where graphical elements are presented together with ranked search results from a search engine, to indicate a degree of association between an author of a document and a topic. In various embodiments, a measure of an author's affinity to a topic is predicted using a machine learning system that takes as input author features derived from documents of the author. In various examples topical affinity scores or labels are used to generate badges and/or to index documents at an index of a search engine. For example, badges may be presented with ranked search results, in pop up displays or in other ways. In examples author features relating to any of recency, frequency and presentation of an author's documents are generated, where presentation features may relate to biographical information such as experience, qualifications, and previous documents.

Topical affinity badges in information retrieval

A recommendation system may use a network of relationships between many different entities to find search results and establish a relevance value for the search results. The relevance value may be calculated by analyzing trust and similarity components of each relationship between the search user and the entity providing the search results. The entities may be, for example, persons associated within express or implied social networks, or corporations or other organizations with a historical or other reputation. The relationships may be created through many different contact mechanisms and may be unidirectional, asymmetric bidirectional, or symmetric bidirectional relationships. The relationships may be different based on topic or other factors.

Hybrid recommendation system

Systems, methods, and computer storage media having computer-executable instructions embodied thereon that facilitate advertisement selection using advancement bids. In embodiments, advertisement attributes in association with an advertisement are referenced. Such advertisement attributes might include an impression bid and a click bid. For a particular advertisement, it is determined if a satisfactory click rate is available. If it is determined that the satisfactory click rate is unavailable for the advertisement, an impression bid is utilized to compete with other advertisements in an advertisement-selection auction. On the other hand, if it is determined that the satisfactory click rate is available for the advertisement, the satisfactory click rate and the click bid are utilized (e.g., via an effective impression value) to compete with the other advertisements in the advertisement-selection auction.

Facilitating advertisement selection using advancement bids

Aspects of the subject matter described herein relate to tuning detection components of a security system. In aspects, a history of alerts is collected. This history is then used together with knowledge about tunable objects of the system to determine parameters of the tunable objects that can be changed to improve detection of the system. Parameters of tunable objects are adjusted in a simulator that determines an effect on alerts in the history based on the adjusted parameters. A recommendation of one or more tuning actions may be provided together with information regarding the effect of each tuning action.

Detection accuracy tuning for security

Security risk for a single IT asset and/or a set of IT assets in a network such as an enterprise or corporate network may be estimated and represented in a visual form by categorizing risk into different discrete levels. The IT assets may include both computing devices and users. The risk categorization uses a security assessment of an IT asset that is generated to indicate the type of security problem encountered, the severity of the problem, and the fidelity of the assessment. The asset value of an IT asset to the enterprise is also assigned. Security risk is then categorized (and a numeric risk value provided) for each IT asset for different problem types by considering the IT asset value along with the severity and fidelity of the security assessment. The security risk for the enterprise is estimated using the numeric risk value and then displayed in visual form.

Estimating and visualizing security risk in information technology systems

Methods, systems, and computer-readable media for search query extension are disclosed. A particular method retrieves a set of queries related to a search query from a query repository. The search query includes one or more search words and at least one search word is included in each query in the set of queries. A set of context words based on one or more computer-readable files accessible to a computer is generated and an intersection is calculated. The intersection is based on the set of queries and the set of context words. An extended search query is generated based on the calculated intersection and the search query.

Search query extension

A honeypot in a computer network is configured for use with a wide variety of computing resources that are defined by a network administrator or user which may include desktop and network resources such as address book contacts, instant messaging contacts, active directory user accounts, IP addresses, and files that contain particular content or that are stored in particular locations. The resources may be real for which protection against leakage is desired, or fake to operate as bait to lure and detect malicious attacks. The honeypot is implemented in an extensible manner so that virtually any resource may be honeypotted to apply honeypot benefits to resources beyond static IP addresses in order to improve both the breadth of information leakage prevention and the detection of malicious attacks.

Personalized honeypot for detecting information leaks and security breaches

Information is automatically located which is relevant to source content that a user is viewing on a user interface without requiring the user to perform an additional search or navigate links of the source content. The source content can be, e.g., a web page or a document from a word processing or email application. The relevant information can include images, videos, web pages, maps or other location-based information, people-based information and special services which aggregate different types of information. Related content is located by analyzing textual content, user behavior and connectivity relative to the source. The related content is scored for similarity to the source. Content which is sufficiently similar but not too similar is selected. Similar related content is grouped to select representative results. The selected content is filtering in multiple stages based on attribute priorities to avoid unnecessary processing of content which is filtered out an early stage.

Growing community of inventors

Herzeliya, Israel

Ron Jack Karidi