International Business Machines Corporation

Karen Wolfe Brannon

Tanveer Fathima Syeda-Mahmood

David James Beymer

Sangeeta T Doraiswamy

Hui-I Hsiao

Ying Chen

Arnon Amir

Inderpal Singh Narang

Suparna Bhattacharya

David Allan Pease

Mark Andrew Smith

Ritwik K Kumar

Ryan John Minniear

Wenling Cai

Ting Chen

Daniel Gruhl

Varun Bhagwan

Steven R Welch

Kenneth Carlin Nelson

Tawei Hu

Neal Ryan Lewis

David Mun-Hien Choy

Gerald Edward Kozina

Lily Liang

Huong Thu Morris

Edward Joseph Gallagher

Colin Compas

Donald Edward Benson

Lesley Mbogo

Matthew Laue

Moritz A W Hardt

Gerald R Clarke

Deb Sudipta

Use of medical workflows where a first medical workflow is obtained from a plurality of medical acts performed in sequence that related to care of a patient. A set of condition-indication rules is applied to the first medical workflow to determine first condition information. The first condition information relates to a likelihood that a first medical condition exists in the patient.

Prediction of diseases based on analysis of medical exam and/or test workflow

Methods and arrangements for handling encrypted messages are disclosed. The method comprising: generating a multi-part encrypted support message responsive to a request therefor; the generating comprising: inputting a base message; identifying at least one span of sensitive information; visually suppressing the at least one span of sensitive information; replacing the at least one span of sensitive information with a redaction notation; and appending at least one redaction message for the at least one span of sensitive information, the at least one redaction message containing data redacted from the at least one span of insensitive information and being encrypted for subsequent opening via at least one key. Other embodiments are disclosed.

Multi-part encrypted messages for support of sensitive systems

Machine learning solutions compensate for data missing from input (training) data and thereby arrive at a predictive model that is based upon, and consistent with, the training data. The predictive model can be generated within a learning algorithm framework by transforming the training data to generate modality or similarity kernels. Similarity values can be generated for these missing similarity values.

Machine learning with incomplete data sets

Fast restoration of referential integrity between metadata and data after they were restored to some inconsistent backup copies in content management archival solutions. An inferred logging mechanism uses separate metadata and data logs to capture recent update activities during normal system conditions with additional object reference information using a method called reference tagging. This requires no system internal changes and introduces no performance overhead. The information in the logs facilitates quick identification of potential referential inconsistencies and allows referential integrity between metadata and data to be restored in a fraction of the time when compared to exhaustive data scans.

Guaranteeing data and metadata referential integrity in content management archival solutions

Systems and methods providing automated extraction of information contained in video data and uses thereof are described. In particular, systems and associated methods are described that provide techniques for extracting data embedded in video, for example measurement-value pairs of medical videos, for use in a variety of applications, for example video indexing, searching and decision support applications.

Systems and methods for automated extraction of measurement information in medical videos

Processing archive content based on hierarchical classification levels is performed by indexing a first newly archived file of a current set of archive content at a first classification level defining a plurality of types of metadata to be derived, wherein the first classification level is selected from a hierarchy of classification levels having at least two classification levels, indexing a second newly archived file of the current set of archive content at a second classification level, wherein the second classification level includes fewer types of metadata than the first class of metadata, and flagging the second newly archived file of the current set of archive content. By classifying and flagging the second newly archived file differently than the first newly archived file, the system can easily retrieve such data at a later instance to update and/or change the data, re-index the data, and the like.

Processing archive content based on hierarchical classification levels

For information lifecyle management (ILM) a process receives a user-definition of ILM actions, data state transitions, and associations among the respective data state transitions and ILM actions and responsively stores a data structure representing the received data state transitions, ILM actions, and correspondence among the respective data state transitions and ILM actions. A process for an ILM model detects a workflow definition event in a monitored process running on a computer system. An information lifecycle flow data structure for the workflow definition is stored responsive to the workflow definition in the monitored process having a data state transition matching one of the user-defined data state transitions, wherein the information lifecycle flow data structure models the matching data state transition, so that the ILM action corresponding to the matching data state transition may be performed for the data responsive to occurrence of a workflow process event in which the matching data state transition occurs.

Business-semantic-aware information lifecycle management

A content management system integrated with a web caching proxy that delivers content according to access control rules. An access control token is generated using a secret key when a user selects a desired object (to be provided only upon token validation, thus the object retrieval and delivery task can be delegated). However, tokens for other content and/or for other users could be generated by a rogue resource manager. If the desired object is already cached, the proxy asks the resource manager to validate the token and then selectively provides the desired object without contacting a library server. Alternately, the proxy itself performs the token validation, but must coordinate with the library server to ensure it has the latest secret key. Finally, the tokens may contain digital signatures generated with a private key and validated with a corresponding public key, so that private keys need not be distributed.

System and method for regulating access to objects in a content repository

A method, an apparatus, a system, a computer program product, and a computer program are disclosed for maintaining consistency of object content () and metadata () related to the object () in a loose transaction model, preferably using SQL Mediated Object Manipulation (SMOM), for object and meta-data updates. The related meta-data () and a reference to the object () are stored in a table of a database. The object is stored externally to the database in an object store. The reference is used to obtain a handle for directly accessing or manipulating the external object. A version number embedded in the handle is then obtained. The embedded version number is then compared with a version number of a latest committed version of the externally stored object to determine if the handle refers to a current version of the externally stored object. Next, the last modification timestamp of the file is compared with the last modification timestamp of the latest committed version, in order to detect uncommitted updates. A mismatch indicates that stale data is being referenced, and in that situation an appropriate error is returned.

Method of maintaining data consistency in a loose transaction model

Managing a content management system. The content management system is one that is configured and controlled to establish a connection between a client and a library server, generate a transaction identifier and insert, a record for the transaction in a tracking table associated with the library server, pass transaction data from the client to a resource manager, process the transaction at the resource manager and record transaction data in a tracking table associated with the resource manager. The resource manager returns transaction success/failure data to the client, compares activity recorded in the tracking tables, and takes corrective action based upon the activity comparison.

Method, system, and program product for transaction management in a distributed content management application

A client-server computing system is disclosed for supporting relational database records and linked external files. The system comprises one or more servers for storing external files, and a plurality of clients interlinked by a network. Each of the servers and certain of the clients have respective filter layers for controlling servicing of application requests relating to external files, for any application command of a protected class relating to an external file made to a client having a filter layer, the respective client filter layer determines the presence of a valid user token associated with the command, then passes the command to a server across the network. The server filter layer, if it recognizes the command, causes the command to be serviced by the server. However, for any application command of a protected class relating to an external file made to a client not having a filter layer, the command will be forwarded across the network to the server filter layer of a server holding the external file and the server filter layer will block the command from being serviced.

Client-server filter computing system supporting relational database records and linked external files operable for distributed file system

Growing community of inventors

Palo Alto, CA, United States of America

Karen Wolfe Brannon