International Business Machines Corporation

Nortel Networks Corporation

Bay Networks, Inc.

Alliant Computer Systems Corporation

Heather D Achilles

Steven R Willis

Eugene Kuznetsov

Paul S Gallo

Nicholas D Matsakis

Charles Robert Morgan

Nirmit V Desai

Shahrokh Daijavad

David Zachary Maze

Wendy Chong

Terry Pearson

Patrick R McManus

Takashi Sakairi

Nanying Yin

Edward S Harriman

Steven Edward Millman

Jan-Christian Nelson

Kenneth R Ballou

Carmelo I Uria

Dan Sullivan

Mike Fortuna

Jim Meredith

Dinesh C Verma

Giora Biran

Jeffrey M Peters

Michael L Ziegler

Amit Golander

Nancy Anne Greco

Catherine Helen Crawford

Bijan Davari

Joseph H Allen

Ai Yoshino

David J Hoeweler

Peter Pothier

Larry B Manor

Edward S Harriman, Jr

Tom Colley

Robert L Fredieu

Joe Latone

Peter F Mycue

Adam E Grant

Randy M Fussell

Christopher P Huff

Kelvin Kakugawa

Al Yoshino

Kevin Klapper

Mark Podlipec

Bill Miller

Kelvin Kakugwa

Personalization with on-device privacy preservation is provided by receiving, at a user device, a generalized recommendation from a remote provider backend, which include a plurality of candidate items responsive to a request from an application running on the user device; generating a plurality of candidate vectors corresponding to the plurality of candidate items representing a corresponding candidate item in a binary format across a plurality of metrics associated with the plurality of candidate items; generating a user vector based on personal data stored on the user device, the user vector representing a user in a binary format across one or more metrics associated with the personal data; determining interaction propensities based on the user vector and the plurality of candidate vectors according to a global model run on the user device; and displaying, on the user device, the plurality of candidate items according to the interaction propensities.

On-device privacy-preservation and personalization

A notification message is received by a communication device The notification message includes a message identifier and a digital signature generated by an original sender of the notification message. That an integrity of message content in the notification message and an original sender of the notification message is verified is determined using the digital signature. Responsive to determining that the notification message is verified, the notification message is forwarded by the communication device using a peer-to-peer communication channel over a mesh network.

Delivering messages to offline devices using peer-to-peer communication

A trained base model is distributed to a set of nodes. From a first node in the set of nodes, a first set of meta-metrics resulting from a transfer learning operation on the trained base model at the first node is collected. The transfer learning at the first node uses first local data available at the first node. The first node is clustered in a cluster with a second node from the set of nodes, in response to a meta-metric in the first set of meta-metrics being within a tolerance value of a corresponding meta-metric in a second set of meta-metrics collected from the second node. A normalized set of model parameters is constructed after an iteration of transfer learning or local learning at the first and second nodes. The normalized set of model parameters is distributed to the first node and the second node in the cluster.

Transfer learning without local data export in multi-node machine learning

Methods and systems for visualizing data include forming hierarchically related graphs having a common time-series axis. The hierarchically related graphs are displayed in a linear arrangement, such that shared values on the common time-series axis align for each graph. The hierarchically related graphs are automatically updated in accordance with a user selection of an element in a data hierarchy by removing graphs below a lowest-order common ancestor in the data hierarchy between the user selection and a previously displayed selection and replacing the removed graphs with new graphs that reflect the user selection.

Visualizing hierarchical time-series data

Methods and systems for visualizing data include forming multiple hierarchically related graphs having a common time-series axis in accordance with a user selection of an element in a data hierarchy. A user interface that includes the plurality of hierarchically related graphs is displayed in a linear arrangement, such that shared values on the common time-series axis align for each graph. Each graph has a parent or child relationship with each adjacent graph.

Aspects of the invention are directed to XML post-processing hardware acceleration for high-speed simple type checking (e.g., string and format checking) and whitespace collapsing. An apparatus for processing text data according to an embodiment includes a post processing engine (PPE); a set of post processing state machines coupled to the PPE; and a memory for storing text data to be processed by the set of post processing state machines; wherein the set of post processing state machines includes a whitespace handler state machine for reading text data from the memory and for applying a whitespace facet to at least 4 bytes of the text data during a clock cycle.

XML post-processing hardware acceleration

A novel and useful system and method of improving the utilization of a special purpose accelerator in a system incorporating a general purpose processor. In some embodiments, the current queue status of the special purpose accelerator is periodically monitored using a background monitoring process/thread and the current queue status is stored in a shared memory. A shim redirection layer added a priori to a library function task determines at runtime and in user space whether to execute the library function task on the special purpose accelerator or the general purpose processor. At runtime, using the shim redirection layer and based on the current queue status, it is determined whether to execute the library function task on the special purpose accelerator or on the general purpose processor.

Utilization of special purpose accelerators using general purpose processors

Systems, methods and apparatus operate to receiving a markup processing program containing a set of markup processing instructions. Each markup processing instruction includes an operation defining a processing operation to be matched to a markup language representation and a corresponding action to be performed upon matching the operation to the markup language representation. The system receives the markup language representation containing encoded items representative of markup language constructs and executes the markup processing program in conjunction with the markup language representation to dynamically control generation of an output control signal indicating how the markup language representation is to be processed for generation of output based on the markup language representation. The markup processing program is applied in real-time to streaming XML data and represents logic instructions that carry out application of post processing such as XPATH expression evaluation, schema validation or other post processing on the stream of XML data.

Method and apparatus for stream based markup language post-processing

A method, computer program product, and system are disclosed for supporting variable data structures and constant data structures in a hardware-based XML PPE. The method may include inserting one or more constant Type-Length-Attributes (TLAs) into an output stream. The constant TLAs are inserted from one or more constant node-sets referenced from a constant table. The method may also include creating one or more variable node-sets. Each variable node-set includes one or more variable TLAs (TLAs saved from a base XML document) and is mapped to a variable table. The method may also include inserting one or more variable TLAs into the output stream inserted from one or more variable node-sets referenced from the variable table.

Supporting variable data structures and constant data structures in a hardware-based extensible markup language (XML) post processing engine (PPE)

Systems, methods and apparatus provide a character process for processing markup language data, such as XML data, by receiving a character stream of markup language data and applying sequences of characters of the character stream to a set of state machines. The set of state machines includes a plurality of construct state machines responsible for processing respective markup language constructs identified by the sequences of characters. The character processor produces, from application of the sequences of characters to the set of state machines, an intermediate representation of the markup language constructs identified by the sequence of characters of the character stream of markup language data. The intermediate representation contains encoded items containing type, length, value representations representative of constructs within the character stream of markup language data.

Methods and apparatus for converting markup language data to an intermediate representation

Systems, methods and apparatus operate to convert a markup language representation such as XML to a data structure format such as a tree containing representations of constructs in the markup language representation. The operations receive a markup language representation containing encoded items representative of markup language constructs and identifying a sequence of encoded items within the intermediate representation that represents a respective markup language construct. In response, the system establishes, within a buffer of nodes, a new node for the sequence of encoded items that represents a respective markup language construct and creates, from the buffer of nodes, a linked data structure that represents relationships between the markup language constructs represented within the markup language representation.

Methods and apparatus for converting a representation of XML and other markup language data to a data structure format

A system assigns identities to received content portions by determining if the content portion has an assigned content identifier, and if so, applying the assigned content identifier to the content portion received. If the content portion does not have an assigned content identifier, the system selects an unassigned content identifier, from the set of unassigned content identifiers, for assignment to the content portion and assigns the unassigned content identifier to the content portion so that the unassigned content identifier becomes an assigned content identifier in a set of assigned content identifiers. The system associates a persistence metric to the assigned content identifier. The persistence metric indicates an importance of longevity of the assignment between the content portion and the content identifier. The system performs a recover operation to unassign content identifiers in the set of assigned content identifier based on a persistence metric of those assigned content identifiers.

Methods and apparatus for assigning content identifiers to content portions

A representation of a markup language data such as XML expressed as a sequence of encoded items provides a data format including a type field containing i) a construct type identifying a type of markup language data construct to which the encoded item corresponds or, ii) a directive type identifying directive information concerning a processing characteristic associated with the sequence of encoded items. If the type field contains a construct type, the encoded item includes i) a length field containing a size of construct data associated with the markup language data construct to which the encoded item corresponds and ii) a value field identifying construct data associated with the markup language data construct to which the encoded item corresponds. For a directive type, the encoded item includes a directive value supplementing the directive information concerning a processing characteristic associated with the sequence of encoded items.

Methods and apparatus for representing markup language data

Growing community of inventors

Hudson, NH, United States of America

Heather D Achilles