International Business Machines Corporation

Massachusetts Institute of Technology

Toyotaro Suzumura

Michiaki Tatsubori

Ilie G Tanase

Mark Richard Nutter

Ching-Yung Lin

Jason L Crawford

Jie Lu

Danny L Yeh

Chun-Fu Chen

Jui-Hsin Lai

Yinglong Xia

Akihiko Tozawa

Tamiya Onodera

Lifeng Nai

Akiko Suzuki

Issei Yoshida

Naohiko Uramoto

Miki M Enoki

Fumiko Satoh

Kaori Fujiwara

Scott Ross Trent

Marco Pistoia

Jinho Lee

Quanfu Fan

Jie Chen

Richard Chen

Charles E Leiserson

Tengfei Ma

Takashi Nerome

Giacomo Domeniconi

Masayuki Ioki

Yuriko Sawatani

Aldo Pareja

Timothy Kaler

Tao B Schardl

Lifeng Nail

Warut Vijitbenjaronk

A system includes a plurality of graph convolutional networks corresponding to a plurality of time steps, each network modelling a graph including nodes and edges, and in turn including a plurality of graph convolution units; an evolving mechanism; and an output layer. Each of the units, for a given one of the time steps, takes as input a graph adjacency matrix, a node feature matrix, and a parameter matrix for a current layer, and outputs a new node feature matrix for a next highest layer. The mechanism takes as input a parameter matrix for a prior time step updates the input parameter matrix, and outputs the parameter matrix for the given time step. The output layer obtains, as input, output of each of the units for a final time step, and based on the output of each of the units for the final time step, outputs a graph solution.

Evolving graph convolutional networks for dynamic graphs

A novel distributed graph database is provided that is designed for efficient graph data storage and processing on modern computing architectures. In particular a single node graph database and a runtime & communication layer allows for composing a distributed graph database from multiple single node instances.

Distributed graph databases that facilitate streaming data insertion and queries by efficient throughput edge addition

Distributed graph databases that facilitate streaming data insertion and queries by reducing number of messages required to add a new edge by employing asynchronous communication

Distributed graph databases that facilitate streaming data insertion and low latency graph queries

A fine-grained scalable time-versioning support for large-scale property graph databases includes receiving a request, wherein the request includes an entity identifier (ID) and reference time information, and searching entries of a first table using the entity ID as a first key. Matching an entry is performed corresponding to the first key in the first table, determining the entry is valid based at least in part on comparing the reference time information of the request with stored time information for the entry, and responsive to the determination, creating a second key, wherein the second key concatenates an ID associated with the entity ID with the reference time information of the request. Searching a second table is performed having entries based at least in part on the second key, and responsive to searching the second table, retrieving data associated with the second key.

Fine-grained scalable time-versioning support for large-scale property graph databases

A generation apparatus that generates a mapping between individual properties included in an object in a program and individual elements of a structured document. The generation apparatus includes: an object tree generation unit that generates a tree structure representing hierarchical structure of the object by assigning the individual properties included in the object to nodes of the tree structure; and a selection unit that selects a mapping minimizing conversion cost of converting the tree structure of the object to a tree structure that includes the individual elements of the structured document as its nodes. The selection is from mappings that associate the individual properties included in the object with the individual elements of the structured document.

Generation apparatus, program, and generation method

A Resilient Distributed Dataset (RDD) processing method, system, and computer program product include sending a different portion including a global identifier and different values of a message from a receiver to a sender and at a receiver side, an already-processed message is extracted from an RDD object repository corresponding to the global identifier of the different portion and the incoming differential portion is applied to the already-processed message.

Differential processing mechanism for spark-based graph computing background

A remote control system detects a directional signal at an alerting sensor package; activates an appliance to accept a control signal in response to the directional signal; accepts the control signal at the appliance from a muscle sensor package; and adjusts operation of the appliance in response to the control signal.

Remote control with muscle sensor and alerting sensor

Techniques facilitating generation of a fused kernel that can approximate a full kernel of a convolutional neural network are provided. In one example, a computer-implemented method comprises determining a first pattern of samples of a first sample matrix and a second pattern of samples of a second sample matrix. The first sample matrix can be representative of a sparse kernel, and the second sample matrix can be representative of a complementary kernel. The first pattern and second pattern can be complementary to one another. The computer-implemented method also comprises generating a fused kernel based on a combination of features of the sparse kernel and features of the complementary kernel that are combined according to a fusing approach and training the fused kernel.

Fusing sparse kernels to approximate a full kernel of a convolutional neural network

Methods and systems for estimating a probability of re-sharing information include extracting keywords from a set of documents addressed to a user. The keywords from the set of documents are weighted according a metric for the user's interest in the keywords' respective source documents to create an interest model. A new document having one or more keywords is received. A likelihood that the user will re-share the new document is determined. The likelihood is based on the interest model and the one or more keywords present in the new document. The new document is automatically responded to based on the determined likelihood.

Estimating probability of spreading information by users on micro-weblogs

At a mobile electronic device, first sensor data is obtained from at least one wearable muscle sensor, suggesting that location tracking will imminently be desirable. A first indication that an external locator signal is not usable is also obtained. Responsive to the first sensor data and the indication, second sensor data, from at least one wearable muscle sensor, indicative of motion of a subject wearing the at least one muscle sensor, is recorded. A second indication that the external locator signal is now usable is obtained. Responsive to the second indication, the external locator signal is used to determine a current location of the mobile electronic device; and the second sensor data is sued to backtrack from the current location of the mobile electronic device to indicate a location of the mobile electronic device at a time when the recording of the second sensor data was initiated.

Trace/trajectory reconstruction via wearable and/or mobile sensors for indoor/outdoor location

Growing community of inventors

Tokyo, Japan

Toyotaro Suzumura