Zengyan Te Zhang

US Patent:

8019899, Sep 13, 2011

Filed:

Aug 28, 2008

Appl. No.:

12/200529

Inventors:

Zengyan Zhang - San Jose CA, US
George E. Mavromatis - Mountain View CA, US
Jinhui Liu - Sunnyvale CA, US

Assignee:

Yahoo! Inc. - Sunnyvale CA

International Classification:

G06F 15/16

US Classification:

709248, 709227, 370230, 718100

Abstract:

Methods, apparatuses and systems directed to processing client requests in distributed systems. Particular implementations of the invention are directed to allowing application servers to return partially processed results after having consumed a time slice allocation based on performance estimates, network performance metrics, and/or system performance metrics.

US Patent:

20060248456, Nov 2, 2006

Filed:

May 2, 2005

Appl. No.:

10/908215

Inventors:

Todd Bender - San Jose CA, US
Keiko Kurita - Los Gatos CA, US
Tram Nguyen - San Jose CA, US
C. Niblack - San Jose CA, US
Zengyan Zhang - San Jose CA, US

Assignee:

IBM CORPORATION - Armonk NY

International Classification:

G06F 17/24

US Classification:

715531000

Abstract:

The present invention provides a method and system of assigning a publication date for at least one electronic document, where the publication date includes the year that the document was published, the month that the document was published, and the day that the document was published. In an exemplary embodiment, the method and system include (1) recognizing the publication date in the document by regular expression pattern matching, (2) if the publication date is ambiguous, resolving the ambiguous publication date, and (3) validating the publication date. In an exemplary embodiment, the recognizing includes determining at least one candidate publication date from the document identifier of the document. In an exemplary embodiment, the recognizing includes determining the publication date from the textual content of the document. In an exemplary embodiment, the recognizing includes determining the publication date from the metadata of the document.

US Patent:

20070016580, Jan 18, 2007

Filed:

Jul 15, 2005

Appl. No.:

11/160943

Inventors:

John Mann - Richmond CA, US
Tram Nguyen - San Jose CA, US
Carlton Niblack - San Jose CA, US
Zengyan Zhang - San Jose CA, US

Assignee:

INTERNATIONAL BUSINESS MACHINES CORPORATION - Armonk NY

International Classification:

G06F 17/30

US Classification:

707006000

Abstract:

The present invention provides a method and system of extracting information about references to entities from a plurality of electronic documents. In an exemplary embodiment, the method and system include () applying at least one document quality measure to each of the plurality of electronic documents, () recognizing the references to entities in the plurality of electronic documents, () using at least one reference quality measure for each of the references to entities, () computing at least one topical category associated with each of the references to entities, () finding at least one co-occurring term associated with each of the references to entities, and () characterizing each of the references to entities by at least one characteristic category.

US Patent:

20070067157, Mar 22, 2007

Filed:

Sep 22, 2005

Appl. No.:

11/234667

Inventors:

Vinay Kaku - Fremont CA, US
Keiko Kurita - Los Gatos CA, US
Carlton Niblack - San Jose CA, US
Jasmine Novak - Mountain View CA, US
Zengyan Zhang - San Jose CA, US

International Classification:

G06F 17/21

US Classification:

704010000

Abstract:

A phrase extraction system combines a dictionary method, a statistical/heuristic approach, and a set of pruning steps to extract frequently occurring and interesting phrases from a corpus. The system finds the “top k” phrases in a corpus, where k is an adjustable parameter. For a time-varying corpus, the system uses historical statistics to extract new and increasingly frequent phrases. The system finds interesting phrases that occur near a set of user-designated phrases. The system uses these designated phrases as anchor phrases to identify phrases that occur near the anchor phrases. The system finds frequently occurring and interesting phrases in a time-varying corpus is changing in time, as in finding frequent phrases in an on-going, long term document feed or continuous, regular web crawl.

US Patent:

20070266016, Nov 15, 2007

Filed:

May 11, 2006

Appl. No.:

11/432265

Inventors:

Scott Holmes - Morgan Hill CA, US
Hongcheng Mi - Sunnyvale CA, US
Sumit Negi - New Delhi, IN
Zengyan Zhang - San Jose CA, US

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

G06F 17/30

US Classification:

707004000

Abstract:

A selection system and method. The selection method comprises receiving, by a computing system, a taxonomy of data related to a specified domain of knowledge on the web. A taxonomy tree is constructed from the taxonomy. A sub tree related to a sub-domain from specified domain is selected from the taxonomy tree. A first list comprising user expected universal resource locators (URLs) related to the sub-domain is received. A second list comprising topic expressions defining each node of the taxonomy sub-tree is generated. A query based on the second list is generated. The query is applied on an index of URLs generated from a web crawling process to generate a third list. A recall value is calculated based on the first list and the third list.

US Patent:

20080033587, Feb 7, 2008

Filed:

Aug 3, 2006

Appl. No.:

11/462100

Inventors:

Keiko Kurita - Los Gatos CA, US
John K. Mann - Richmond CA, US
Ross Nelson - Sunnyvale CA, US
Tram T. Nguyen - San Jose CA, US
Carlton W. Niblack - San Jose CA, US
Zengyan Zhang - San Jose CA, US

International Classification:

G06F 19/00

US Classification:

700100

Abstract:

Disclosed are embodiments of a method of mining data and a method of evaluating that data in order to discover significant changes in conditions (e.g., changes in activities, events, associations, affiliations, market preferences, etc.). The data mining technique uses predetermined scenarios that characterize specific changes as well as key variables that are relevant to those scenarios. These variables are input as mining parameters into a data mining tool. Retrieved data is analyzed and the results are evaluated. One technique of evaluating the results includes displaying them in a visual format (e.g., graphs, tables) along with additional information (e.g., lists of documents or portions of documents containing data relevant to the displayed results). A user evaluates the displayed results and additional information in order to identify data that should be filtered, to identify trends and/or patterns in the data, and to assess the trends and/or patterns.

US Patent:

20100082608, Apr 1, 2010

Filed:

Sep 30, 2008

Appl. No.:

12/242273

Inventors:

Mauricio Mediano - Cupertino CA, US
Jason Zien - Mountain View CA, US
Zengyan Zhang - San Jose CA, US
Jinhui Liu - Sunnyvale CA, US
Song Lin - Sunnyvale CA, US

Assignee:

YAHOO! INC. - Sunnyvale CA

International Classification:

G06F 7/06
G06F 17/30

US Classification:

707723, 707E17109

Abstract:

Embodiments of methods, apparatuses, devices and systems associated with determining search results are disclosed.

US Patent:

20100121712, May 13, 2010

Filed:

Nov 12, 2008

Appl. No.:

12/269383

Inventors:

Ben Shahshahani - Menlo Park CA, US
Vanja Josifovski - Los Gatos CA, US
Peiji Chen - Saratoga CA, US
Zengyan Zhang - San Jose CA, US
Jianchang (JC) Mao - San Jose CA, US
Ruofei (Bruce) Zhang - San Jose CA, US
Sanjay Kshetramade - Fremont CA, US

Assignee:

Yahoo! Inc. - Sunnyvale CA

International Classification:

G06F 17/30
G06Q 30/00

US Classification:

705 1454, 705 10

Abstract:

A system for selecting advertisements for a web page. An advertisement request is received and the web page content is evaluated to identify a first advertisement query based on the content of the web page and a second advertisement query based on the content of the web page or a behavioral feature. A first set of advertisements is retrieved based on the first advertisement query and a second set of advertisements is retrieved based on the second advertisement query. The first set of advertisements is blended with the second set of advertisements to form a blended advertisement set. The blended advertisement set is provided for display with the web page.