International Business Machines Corporation

Rensselaer Polytechnic Institute

Micro Focus Lcc

Mohammed J Zaki

Lingfei Wu

Yu Cheng Chen

Rakesh Agrawal

Ruth Bergman

Omer Barkol

Ching-Tien Ho

Shahar Golan

Arik Sityon

Yifat Felder

Pranay Anchuri

Yu Chen

An initial noisy graph topology is obtained and an initial adjacency matrix is generated by a similarity learning component using similarity learning and a similarity metric function. An updated adjacency matrix with node embeddings is produced from the initial adjacency matrix using a graph neural network (GNN). The node embeddings are fed back to revise the similarity learning component. The generating, producing, and feeding back operations are repeated for a plurality of iterations.

Iterative deep graph learning for graph neural networks

For a passage text and a corresponding answer text, perform a word-level soft alignment to obtain contextualized passage embeddings and contextualized answer embeddings, and a hidden level soft alignment on the contextualized passage embeddings and the contextualized answer embeddings to obtain a passage embedding matrix. Construct a passage graph of the passage text based on the passage embedding matrix, and apply a bidirectional gated graph neural network to the passage graph until a final state embedding is determined, during which intermediate node embeddings are fused from both incoming and outgoing edges. Obtain a graph-level embedding from the final state embedding, and decode the final state embedding to generate an output sequence word-by-word. Train a machine learning model to generate at least one question corresponding to the passage text and the answer text, by evaluating the output sequence with a hybrid evaluator combining cross-entropy evaluation and reinforcement learning evaluation.

Natural question generation via reinforcement learning based graph-to-sequence model

Aspects described herein include a method of conversational machine reading comprehension, as well as an associated system and computer program product. The method comprises receiving a plurality of questions relating to a context, and generating a sequence of context graphs. Each of the context graphs includes encoded representations of: (i) the context, (ii) a respective question of the plurality of questions, and (iii) a respective conversation history reflecting: (a) one or more previous questions relative to the respective question, and (b) one or more previous answers to the one or more previous questions. The method further comprises identifying, using at least one graph neural network, one or more temporal dependencies between adjacent context graphs of the sequence. The method further comprises predicting, based at least on the one or more temporal dependencies, an answer for a first question of the plurality of questions.

Conversation history within conversational machine reading comprehension

A method for discovering representative composite configuration item (CI) patterns in an IT system that includes a plurality of configuration items may include data mining a graph representing the IT system to extract extended frequent composite CI patterns. The method may also include clustering the extended frequent composite CI patterns into clusters based on similarity between the maximal frequent composite CI patterns. The method may further include extracting a representative composite CI pattern for each of the clusters, and using an output device, outputting the representative composite CI pattern for each of the clusters.

Discovering representative composite CI patterns in an it system

A method and system for generating a decision-tree classifier in parallel in a shared-memory multiprocessor system is disclosed. The processors first generate in the shared memory an attribute list for each record attribute. Each attribute list is assigned to a processor. The processors independently determine the best splits for their respective assigned lists, and cooperatively determine a global best split for all attribute lists. The attribute lists are reassigned to the processors and split according to the global best split into the lists for child nodes. The split attribute lists are again assigned to the processors and the process is repeated for each new child node until each attribute list for the new child nodes includes only tuples of the same record class or a fixed number of tuples.

Parallel classification for data mining in a shared-memory multiprocessor system

Growing community of inventors

Troy, NY, United States of America

Mohammed J Zaki