Microsoft Technology Licensing, LLC

Galen Andrew

Jianfeng Gao

Robert L Rounthwaite

Emre Mehmet Kiciman

Silviu-Petru Cucerzan

Kristina Nikolova Toutanova

Jamie Paul Buckley

Xiaoxin Yin

Mark Edward Johnson

Joanna Chan

Sooho Park

A system described herein includes a receiver component that receives a dataset that is stored in a computer-readable medium of a computing device, wherein the dataset includes a plurality of queries issued by users to a search engine and a plurality of search results selected by the users upon issuing the plurality of queries. A distribution determiner component determines click distributions over the search results selected by the users with respect to the plurality of queries. A labeler component labels at least two queries in the plurality of queries as being substantially similar to one another based at least in part upon the click distributions over the search results selected by the users with respect to the plurality of queries.

Determining a similarity measure between queries

Described herein is a system that facilitates assigning indications of usefulness to query suggestions. The system includes a query suggestion generator component that receives a query and generates a query suggestion based at least in part upon the received query. A model component outputs an indication of usefulness with respect to the query suggestion, wherein the model component is a machine-learned model of user behavior with respect to query suggestions.

Query suggestion generation

An algorithm that employs modified methods developed for optimizing differential functions but which can also handle the special non-differentiabilities that occur with the L-regularization. The algorithm is a modification of the L-BFGS (limited-memory Broyden-Fletcher-Goldfarb-Shanno) quasi-Newton algorithm, but which can now handle the discontinuity of the gradient using a procedure that chooses a search direction at each iteration and modifies the line search procedure. The algorithm includes an iterative optimization procedure where each iteration approximately minimizes the objective over a constrained region of the space on which the objective is differentiable (in the case of L-regularization, a given orthant), models the second-order behavior of the objective by considering the loss component alone, using a 'line-search' at each iteration that projects search points back onto the chosen orthant, and determines when to stop the line search.

Limited-memory quasi-newton optimization algorithm for L1-regularized objectives

Systems and methods for selecting a ranker for statistical natural language processing are provided. One disclosed system includes a computer program configured to be executed on a computing device, the computer program comprising a data store including reference performance data for a plurality of candidate rankers, the reference performance data being calculated based on a processing of test data by each of the plurality of candidate rankers. The system may further include a ranker selector configured to receive a statistical natural language processing task and a performance target, and determine a selected ranker from the plurality of candidate rankers based on the statistical natural language processing task, the performance target, and the reference performance data.

Ranker selection for statistical natural language processing

Growing community of inventors

Redmond, WA, United States of America

Galen Andrew