Robert T Farrell

US Patent:

7281002, Oct 9, 2007

Filed:

Mar 1, 2004

Appl. No.:

10/790642

Inventors:

Robert G. Farrell - Cornwall NY, US

Assignee:

International Business Machine Corporation - Armonk NY

International Classification:

G06F 17/30

US Classification:

707 3, 707 5, 707 7

Abstract:

The present invention provides system, methods, and apparatus for organizing objects obtained from data retrieval. It provides a method to assemble the objects into a set based upon their relevance ranking, associated metadata including categories, and the proximity of these categories in a graph. The method provides for selecting objects according to estimated difficulty, resource type, media format, or other criteria. It also provides for connecting objects into a best path in the graph. It also provides for selecting objects to fit a desired total duration and coverage of relevant categories in the graph. It also provides for organizing the objects into a particular sequence according to the particular metadata. The system, methods, and apparatus are suitable for use in information systems, learning management systems, and other systems employing data retrieval.

US Patent:

7752199, Jul 6, 2010

Filed:

Aug 29, 2007

Appl. No.:

11/847221

Inventors:

Robert G. Farrell - Cornwall NY, US

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

G06F 17/30

US Classification:

707723, 707789, 707955

Abstract:

The present invention provides system, methods, and apparatus for organizing objects obtained from data retrieval. It provides a method to assemble the objects into a set based upon their relevance ranking, associated metadata including categories, and the proximity of these categories in a graph. The method provides for selecting objects according to estimated difficulty, resource type, media format, or other criteria. It also provides for connecting objects into a best path in the graph. It also provides for selecting objects to fit a desired total duration and coverage of relevant categories in the graph. It also provides for organizing the objects into a particular sequence according to the particular metadata. The system, methods, and apparatus are suitable for use in information systems, learning management systems, and other systems employing data retrieval.

US Patent:

7805440, Sep 28, 2010

Filed:

Apr 5, 2002

Appl. No.:

10/496778

Inventors:

Robert G. Farrell - Cornwall NY, US
Douglas N. Gordin - Katonah NY, US

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

G06F 7/00
G06F 17/30

US Classification:

707728

Abstract:

The system has a collection of a plurality of objects. Each object defines a node in a k-partite graph, such that, the nodes can be divided into a number of mutually exclusive sets such that all of the nodes are in exactly one of the sets; further edges occur only between nodes in different sets; The system also has a simplification process that aggregates one or more of the nodes into one or more categories and identifies a category node corresponding to each category. The category node inherits the mode and the edges of all the nodes in the respective category. Further, the system contains Directed Acyclic Graphs Indices (DAGIs) whose nodes may have a 1-1 mapping with the nodes in the k-partite graph. These indices can be used to aggregate and hide nodes in the k-partite graph. Aggregation occurs by selecting one or more non-leaf nodes in the DAGI and aggregating all descendent nodes.

US Patent:

8019858, Sep 13, 2011

Filed:

Sep 22, 2008

Appl. No.:

12/234789

Inventors:

Frederic Bauchot - Saint-Jeannet, FR
Flemming Boegelund - Frederikssund, DK
Robert George Farrell - Cornwall NY, US
Gerard Marmigere - Drap, FR

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

G06F 15/173

US Classification:

709224, 709232

Abstract:

A method for utilizing system lag to send facts to an end user. The method begins by monitoring a computer system to identify said system lag. After identifying system lag, the method determines an estimated length of the system lag, the estimated length being a measurement of time. The estimated length takes into account previously identified system lag lengths to better approximate the current system lag length. After determining the estimated length of the system lag, the method sends a fact to the end user during the system lag. The fact sent to the end user has a length equal to or less than the estimated length of the system lag. Finally, the method stores the actual length of the system lag in order to improve subsequent determinations of the estimated length.

US Patent:

20090006173, Jan 1, 2009

Filed:

Jun 29, 2007

Appl. No.:

11/771387

Inventors:

Robert George Farrell - Cornwall NY, US
Jianying Hu - Bronx NY, US
Sarah Campbell McAllister - Baton Rouge LA, US
Aleksandra Mojsilovic - New York NY, US
Bonnie Kathryn Ray - Nyack NY, US

Assignee:

INTERNATIONAL BUSINESS MACHINES CORPORATION - Armonk NY

International Classification:

G06Q 10/00

US Classification:

705 9

Abstract:

A method for identifying and using historical work patterns to build high-performance project teams, in one aspect, may comprise identifying historical data associated with one or more past projects, determining from said historical data, one or more patterns in team member attributes that are correlated with at least one of an individual determined to be successful and a project determined to be successful, and generating one or more staffing plans based on said determined patterns. A system and program storage device for performing finctionalities of the method are also provided.

US Patent:

20220391733, Dec 8, 2022

Filed:

Aug 16, 2022

Appl. No.:

17/888565

Inventors:

- Armonk NY, US
Robert G. Farrell - Cornwall NY, US

International Classification:

G06N 5/04

Abstract:

A policy violation detection computer-implemented method, system, and computer program product, includes predicting a violation of a portion of a policy by a future action and delivering an alert of the violation to a user device, alert settings for privacy are toggled on the user device based on a user profile.

US Patent:

20210279422, Sep 9, 2021

Filed:

May 18, 2021

Appl. No.:

17/323584

Inventors:

- Armonk NY, US
Robert G. Farrell - Cornwall NY, US
Nicolas Rodolfo Fauceglia - New York NY, US
Alfio Massimiliano Gliozzo - Brooklyn NY, US

International Classification:

G06F 40/30
G06F 16/335
G06F 16/901
G06F 40/211
G06F 40/247
G06F 40/268
G06F 40/284
G06F 40/295

Abstract:

Systems and techniques that facilitate spurious relationship filtration from external knowledge graphs based on distributional semantics of an input corpus are provided. In one or more embodiments, a context component can generate a context-based word embedding of one or more first terms in a document collection. The embedding can yield vector representations of the one or more first terms. The one or more first terms can correspond to knowledge terms in one or more first nodes of a knowledge graph. In one or more embodiments, a filtering component can filter out a relationship between the one or more first nodes and a second node of the knowledge graph based on a similarity value being less than a threshold. The similarity value can be a function of the vector representations of the one or more first terms. In various embodiments, cosine similarity can be used to compute the similarity value.

US Patent:

20210125058, Apr 29, 2021

Filed:

Oct 29, 2019

Appl. No.:

16/666800

Inventors:

- Armonk NY, US
Robert G. Farrell - Cornwall NY, US
Nicholas Brady Garvan Monath - Northhampton MA, US
Michael Robert Glass - Bayonne NJ, US
Md Arafat Sultan - Croton-on-Hudson NY, US

International Classification:

G06N 3/08
G06N 5/04
G06F 17/27
G06K 9/62

Abstract:

Training a machine learning model such as a neural network, which can automatically extract a hypernym from unstructured data, is disclosed. A preliminary candidate list of hyponym-hypernym pairs can be parsed from the corpus. A preliminary super-term-sub-term glossary can be generated from the corpus, the preliminary super-term-sub-term glossary containing one or more super-term-sub-term pairs. A super-term-sub-term pair can be filtered from the preliminary super-term-sub-term glossary, responsive to detecting that the super-term-sub-term pair is not a candidate for hyponym-hypernym pair, to generate a final super-term-sub-term glossary. The preliminary candidate list of hyponym-hypernym pairs and the final super-term-sub-term glossary can be combined to generate a final list of hyponym-hypernym pairs. An artificial neural network can be trained using the final list of hyponym-hypernym pairs as a training data set, the artificial neural network trained to identify a hypernym given new text data.