Maxtor Corporation

Microsoft Technology Licensing, LLC

Seagate Technology Incorporated

Linkedin Corporation

Lin Guo

Yu Sun

Don Brunnett

Xiaoping Hu

Wei Guo

Yunfeng Li

David Marcus Tung

Stan Shepherd

S Craig Smith

Erhard Theodor Schreck

Yuan Chen

Youping Deng

Xin Yang

Y-j Dennis Chen

Brian P Rigney

Bo Long

Shakti Dhirendraji Sinha

Ganesh Venkataraman

Stanley H Shepherd

Viet Thuc Ha

Huiji Gao

Asif Mansoor Ali Makhani

Jianling Zhong

Senthil Sundaram

Weiwei Guo

Jye-Kai Chang

Bruce A Liikanen

Charles R Watt

Thao Anh Nguyen

Yih-Jen Dennis Chen

Junaid Ahmed

Dhruv Arya

John Charles Purkett

John Edward Scura

Wentao Yan

Abhishek Kumar Singh

Richard Oswald

Bill S Wang

John VanLaanen

John W Vanlaanen

Xiaokun L Chew

Walter Sun

Jason Adler

Kok-Meng Lee

Jong-Ming Lin

Mark Heimbaugh

Saurabh Kataria

Oliver W Northrup

Ada Cheuk Ying Yu

Mark S Seymour

Abhimanyu Lad

William Flynn

Yinzhe Yu

Jay Goyal

Chris Hawks

Jingfeng Li

Ramesh Dommeti

Vinay Deshpande

Dave McMurtrey

Tim Glassburn

Matthew A O'Hara

Todd Franks

Hamed Firooz

Dennis Chen

Tony Hurtado

William M Marvin, Jr

Douglas Martin

Harry Douglas Garner, Jr

Thomas Arthur Ledbetter

Brahm Kiran Singh

Paul Yeagle

David Hyunchul Shim

Jong-Min Lin

Xiao (John) Zhang

Georgie Tan

Heng-Poh Khong

Donald Brunett

Machine learning based method for generating personalized query suggestions is described. Different users may have different search intent even when they are inputting the same search query. The technical problem of personalizing search query suggestions produced by a machine learning model is addressed by extending the sequence to sequence machine learning model framework to be able to take into consideration additional, personalized features of the user, such as, e.g., profile industry, language, geographic location, etc. This methodology includes an offline model training framework as well as an online serving framework.

Generating personalized query suggestions

A neural related query generation approach in a search system uses a neural encoder that reads through a source query to build a query intent vector. The approach then processes the query intent vector through a neural decoder to emit a related query. By doing so, the approach gathers information from the entire source query before generating the related query. As a result, the neural encoder-decoder approach captures long-range dependencies in the source query such as, for example, structural ordering of query keywords. The approach can be used to generate related queries for long-tail source queries, including long-tail source queries never before or not recently submitted to the search system.

Neural related search query generation

Techniques for performing a database search using a rewritten and annotated query are disclosed herein. In example embodiments, a profile lexicon is generated from a set of raw user profiles. A click-through lexicon is generated from a raw query log. A machine-learning model is trained for entity prediction using selected data. Query tagger data is generated using the profile lexicon, the click-through lexicon, and the machine-learning model. A raw query is received. The raw query is rewritten as an annotated query based on the generated query tagger data. A search of a database is performed using the annotated query. Results of the search are returned in response to the receiving of the raw query for presentation in a user interface.

Entity- and string-based search using a dynamic knowledge graph

Disclosed are systems, methods, and non-transitory computer-readable media for expanding search queries. A search system determines a set of candidate alternate search terms based on historical search logs that include records of previously submitted search terms, corresponding search results that were presented to users, and corresponding search results that were selected by the users. The set of candidate alternate search terms is selected from titles of the corresponding search results that were selected by the users. The search system ranks the set of candidate alternate search terms based on determined probabilities that each of the alternate candidate search terms will be selected if presented to a user, and selects a first candidate alternate search term from the set of candidate alternate search terms based on the ranking. The search system generates an expanded search term based on the first candidate alternate search term.

Expanding search queries

This disclosure relates to systems and methods for searching names using name clusters. A method includes receiving names, generating a plurality of phonetic cluster identifiers, forming a plurality of name clusters by grouping the names having an equivalent cluster id, removing names from the respective name clusters that differ from a root name by more than either a particular spelling of a phonetic sound or a specific member's reformulation according to a reformulation dictionary, and suggesting one or more names by generating a phonetic cluster id for the received name using the database of phonetic associations and returning names found in the name cluster that matches the phonetic cluster id.

Search engine using name clustering

A system and method for personalized search user searcher features may include obtaining a search term from a member of a social network at a user device via the network interface. An initial result may be generated based on the search term, including a first group of content items from a social network and stored in a content database, the content items including member profiles of members of the social network. Each of the content items of the first group may be ranked based on information indicative of interactions from an activity database with the content items of the first group, the interactions being by at least a second user of the social network different than the first user. A second group of the content items may be displayed, including at least some of the first group of the content items, based on the rank of the first group of the content items.

Personalized search using searcher features

A system and method for personalized search based on searcher interest may include obtaining a search term from a member of a social network at a user device via the network interface. An initial result may be generated based on the search term, including a first group of content items from a social network and stored in a content database, the content items including member profiles of members of the social network. Each of the content items of the first group may be ranked based on information from an activity database, the activity database storing the information related to the social network, the activities including interactions with search results that include ones of the member profiles. A second group of the content items may be displayed, including at least some of the first group of the content items, based on the rank of the first group of the content items.

Personalized search based on searcher interest

Methods, systems, and computer-storage media for improving the freshness, or the apparent freshness, of search results are described. In an embodiment, the first portion of search results presented on a search results page are based on responsiveness to the search query and a second portion of results describe only recently published documents that are responsive to the search query. In an embodiment, a more recent version of the document, which is not directly used to determine responsiveness, is used to build the caption for a search result. Another way to make search results appear fresh is to include a publication time within the search result caption. In one embodiment, the publication time is generated by calculating a point in time between when a document is first added to a search index and the previous time the search engine visited the site where the document was found.

Enhancing freshness of search results

Various embodiments of the present invention allow for the generation of at least one estimated value of a first parameter that is derived from a regularly measured value of a second parameter. In addition, a single control feedback signal can be adjusted via updating of the estimated and regularly measured values in a control feedback loop. At least one spiral track can be written to storage media based on the single control feedback signal, where the single control feedback signal can be received from an compensating circuit that receives the at least one estimated value of the first parameter, which can refer to position, and the at least one regularly measured value, which can refer to velocity.

Self-writing of servo patterns

Methods and structure for switching heads on a multiple recording surface disk drive that uses bi-directional radial writing toward a zipper zone at a middle diameter. When switching from a first head to a second head, features and aspects hereof assure that the second head will not be positioned in the zipper zone of the corresponding second surface and hence will not lose track following servo synchronization and control. In one aspect, added buffer or guard tracks having valid servo information but no data are added on either side of the zipper zone to assure that under the worst-case misalignment of surfaces the second head will not be positioned in its corresponding zipper zone when switching from the first head.

Methods and structure for improved disk head switching to avoid zipper zones

A disk drive slider positioner () includes a frame () having a pair of arms () that cantilever from a base (). A slider () is positioned within frame (). A pair of slots () extend down through the frame () and also extend along part of the corresponding arm (). Each slot () includes a corresponding first slot section () and a wider corresponding second slot section (). A material () that is more pliable than the frame () preferably occupies the entire length of each first slot section (). Excess first material () may overflow into the corresponding wider second slot section (), thereby reducing the potential for excess first material () occupying any portion of the space between a leading edge () of the slider () and the frame ().

Disk drive microactuator frame with epoxy overflow reservoir

In accordance with various embodiments, a rotatable member is mounted to a spindle hub so that initial servo data previously provided to the rotatable member are eccentrically offset with respect to a rotational center of the spindle hub. Compensation vales are determined to characterize said offset, after which final servo data are provided to the rotatable member in relation to the initial servo data and the determined compensation values. The final servo data are nominally concentric with the rotational center of the spindle hub.

Servo writing with offset compensation for prewritten reference

A control source and drivers for a thermal actuator fly height controlled heads in a disk drive system includes provisions for driving a heater element to dissipate various levels of power, and thus actuate the head by a desired amount, depending on whether the head is in a write mode or a read mode. Further, overshoot and undershoot are achieved as desired by providing initial values and equilibrium values and through the use of a control capacitor to provide a time constant to the drive signal to the heater element.

Method of operating read and write transducers by controlling thermal actuation

In accordance with various embodiments, initial servo data are written to a storage medium as a series of radially overlapped spiral segments. Final servo data are thereafter written to the medium while concurrently servoing on said overlapped segments.

Servo writing using radially overlapped servo segments

The present invention is directed to a method and apparatus for moving off of a disk drive's crash stop in a consistent manner.

Method and apparatus for moving off of a disk drive's crash stop in a consistent manner

Growing community of inventors

Saratoga, CA, United States of America

Lin Guo