Microsoft Technology Licensing, LLC

Twitter, Inc.

Nipoon Malhotra

Nitin Agrawal

Sanaz Ahari

Michael Austin Schechter

Kieran Richard McDonald

Utkarsh Srivastava

Ran Gilad-Bachrach

Peter Bailey

Brian W MacDonald

Mauktik Gandhi

Xin Liu

Ram Viswanathan

Reid Andersen

Vikram Kakumani

Liwei Chen

Sarath Madakasira

Kumaresh Pattabiraman

Christopher A Meyers

James Francis Gilsinan, Iv

Tapas Kanungo

Hyun Duk Kim

James Brian Lawson

Mairéad Áine O'Donovan

Kimberly M Vlcek

Yiming Li

Kiran Sagi

A real-time messaging platform allows advertiser accounts to pay to insert candidate messages into the message streams requested by account holders. To accommodate multiple advertisers, the messaging platform controls an auction process that determines which candidate messages are selected for inclusion in a requested account holder's message stream. Selection is based on a bid for the candidate message, the message stream that is requested, and a variety of other factors that vary depending upon the implementation. The process for selection of candidate messages generally includes the following steps, though any given step may be omitted or combined into another step in a different implementation: targeting, filtering, prediction, ranking, and selection.

Campaign goal setting in a messaging platform

Computer-readable media, computer systems, and computing devices for initiating a queryless search is provided. In this regard, a search can be initiated without a user inputting a query. In one embodiment, the method includes recognizing a search indication provided by a user to indicate a desire to view one or more search results related to content being presented. Such a search indication can be performed without accessing a web page associated with a search engine. In accordance with the search indication, contextual information related to the content being presented can be automatically captured. Thereafter, the captured contextual information is provided to initiate a search for information corresponding with the contextual information. Search results corresponding with the contextual information are received and presented.

Queryless search based on context

Systems, methods, and computer-readable storage media for providing query-completion suggestions configured to be searched against vertical information domains in response to receiving a general web search query are provided. At least a portion of a general web search query is received. As the portion of the general web search query is received, query-completion suggestions are determined as is a likely primary intent associated with at least a portion of the determined query-completion suggestions. If the likely primary intent associated with a particular query-completion suggestion is for information associated with a particular vertical information domain other than the general web domain, the query-completion suggestion is provided such that selection thereof causes searching of the query-completion suggestion against the particular information domain for search results.

Domain specific query-completion suggestions

Computer-readable media, computer systems, and computing devices facilitate providing a comparison experience to a user in response to a search query. Upon receiving a search query from the user, entities are extracted from the query. The entities are associated with entity classes. The entities, entity classes, previous user behavior, and other information are used to infer whether the user likely is engaging in a comparison task. If the inference indicates that the user likely is engaging in a comparison task, a comparison experience is generated and access to the comparison experience is provided to the user.

Providing comparison experiences in response to search queries

Use of software applications is detected by categorizing components of applications into a usage manifest and implementing a usage detection background service on a client PC that monitors the components according to the usage manifest. The application components are categorized based on the mode of user interaction as well as the component's correlation to active use of an application. The background service tracks events and activities associated with the application components to generate usage metrics that include the frequency of unique launches of an application and the duration of each unique use. A usage manager for the background service may utilize the usage metrics for a component independently, or combine metrics for multiple components in cases where applications work in an interactive manner (such as a plug-in to a web browser) in order to compute application usage by comparing the metrics against predefined thresholds.

Reliable and accurate usage detection of a software application

Technologies are described herein for placing search results on a search engine results page (SERP). A query may be received. The query may be transmitted to a plurality of search result providers. A first set of search results and a second set of search results may be received from the search result providers. Intent features may be extracted from the first set of search results. User intent of the second set of search results may be inferred based on the extracted intent features. The first set of search results and the second set of search results may be ranked based on the inferred user intent. The SERP may be rendered according to the ranked first set and second set of search results.

Placement of search results using user intent

Methods and computer-storage media having computer-executable instructions embodied thereon that facilitate generating a machine-learned model for ranking search results using click-based data are provided. Data is referenced from user queries, which may include search results generated by general search engines and vertical search engines. A training set is generated from the search results and click-based judgments are associated with the search results in the training set. Based on click-based judgments, identifiable features are determined from the search results in a training set. Based on determining identifiable features in a training set, a rule set is generated for ranking subsequent search results.

Ranking search results using click-based data

Growing community of inventors

Redmond, WA, United States of America

Nipoon Malhotra