International Business Machines Corporation

Jonathan Limburn

Scott Schumacher

Martin Oberhofer

Barry Neil Dow

Albert Maier

Dan Jeffrey Mandelstein

Alexander T Bentley

Olena Woolf

Manish Anand Bhide

Ivan Matthew Milman

Graham White

Kevin Christopher Brown

Michael Cobbett

Edward G Jellard

Harivansh Kumar

Nicholas Kanellos

Embodiments relate to a system, program product, and method for generating an enhanced feature catalog for a predictive model. The embodiments disclosed herein include capturing predictive model design time information including training data lineage metadata to determine the features of the training data, model design time measurements, and model design time metadata. Once the predictive model is built, the training data lineage metadata is used to capture the features that will be maintained within a feature catalog. The model design time measurements and model design time metadata provide further correlation between the predictive model and the features. Runtime metrics on the predictive model create additional correlations between the captured data and metadata with the features in the feature catalog to expeditiously identify the relevant features of the predictive model.

Feature catalog enhancement through automated feature correlation

Cross-matching of match indexes assigned to respective data sources can be performed. Each match index can store at least a portion of data from a subset of data records in the data source to which the match index is assigned. Such data can be stored in the match index as match index records. Responsive to identifying matches between match records of different match indexes, the data records in the data sources to which the match records correspond can be updated to add links between the respective data records.

Data resolution among disparate data sources

Task handling in a master data management (MDM) system. A task pattern associated with a task completed is identified, in a task handling system. If the identified task pattern matches an existing task pattern in a task patterns database, then a pattern count value of the matching existing task pattern is incremented, and the task patterns database is searched for task patterns that are similar to the matching existing task pattern. If similar task patterns are identified, for each of the similar task patterns, a duplicate pattern count is specified that is based on the number of identified similar task patterns. If the duplicate pattern count for any one of the similar task patterns exceeds a specified threshold, a new task is created comprising data on completion actions for the matching existing task pattern and each of the one or more similar task patterns.

Task handling in a master data management system

User modeling facilitates use of a UML user model to define roles, goals, and tasks inside an organization, and defines how the roles, goals, and tasks relate to business entities within the organization. This information can be used to generate a user interface (UI) that facilitates management of the defined business entities within the organization. Each task defined inside the model, which may be represented as a stereotyped class in the UML class diagram, has a number of explicit steps associated with the task. These steps are outlined in UML activity diagrams and defined in a task interaction pattern. A modular approach to constructing UML activity diagram using the task interacting pattern to reduce the size of the model permits greater flexibility in generating the UI.

Modularizing steps within a UML user model interaction pattern

An ontology is automatically generated for a set of data from multiple data sources. A semantic network of known concepts serves as an ontology template for a target domain of knowledge with known concepts defined as base entity types. Logical groupings of data and associated technical metadata are read from the data sources. Data discovery techniques are applied to detect semantic and/or syntactic classification of data attributes in the logical groupings of data. For each of the logical groupings of data, an entity type with properties for the data attributes are generated and the generated entity type is added to the semantic network with classifications of the properties as derived from the applied data discovery techniques. Semantic meanings of the generated entity types are generated and associated with other entity types within the semantic network and the semantic network is output as a resultant ontology for the set of data.

Automatic ontology generation

Method and system are provided for refactoring of databases to include soft type information. The method may include: carrying out data profiling of soft type data to generate an attribute list with mapping suggestions to a hardened database structure; generating a data model definition and extract, transform, and load logic for transforming the soft type data based on the attribute list and mapping suggestions; executing the data model definition to create a new or modified database structure; moving data from an existing database structure to the new or modified database structure; and regenerating a services interface for access to the data.

Refactoring of databases to include soft type information

Task patterns may exist in completed tasks within a task handling system. A task pattern may comprise task data and metadata and an action that the task handling system uses to complete the task. A task pattern may be identified and compared with task patterns in a database. If the identified task pattern matches a task in the database, a task pattern count for the task in the database can be incremented. If this incrementing results in the task pattern count exceeding a threshold, then the system can search for and identify other active tasks that match the task pattern. The appropriate action for the task pattern can be applied to each of the matched active tasks and each of the active tasks can then be closed.

Task handling in a multisystem environment

A processor implemented method, system, and/or computer program product switches an electronic social presence between a first device and a second device. An electronic signal indicating a signing in to a social presence server by the first device is received. A determination is made as to whether the first device and the second device are physically located farther apart than a predetermined distance. In response to determining that the first device and the second device are physically located farther apart than the predetermined distance, the first device signs out from the social presence server, the second device signs in to the social presence server, and a history associated with the electronic social presence is transferred from the first device to the second device.

Switching of electronic social presence between devices

A tool for performing a computer action based on a user's touch on a touch screen. The user's touch is characterized by certain parameters which are compared to predefined parameters which correspond to the hand that was used to touch the screen and a specific computer action. If the parameters match, within a specific error tolerance, the computer action is performed.

Computer action detection

A server computer receives a user identifier from a user computer web browser program. The server computer identifies one or more uncompleted tasks associated with the user identifier that can be completed from the user computer via the user computer web browser program. The server computer then transmits to the user computer one or more universal resource locators (URLs) associated with the one or more uncompleted task, and the user computer provides access to the web pages referenced by the one or more URLs via the user computer web browser program. In certain embodiments, the server computer identifies the one or more uncompleted tasks by accessing a database that includes uncompleted tasks by user identifier and a URL associated with each uncompleted task. In other embodiments, the server updates the database when an uncompleted task associated with a user identifier is completed.

Browser start pages based on remote content

A method displays a Unified Modeling Language (UML) diagram on a graphic user interface. The UML diagram can have, for example, nodes representing objects of an object-oriented system. Further, input is received regarding the location of a user-directed pointer on the UML diagram. The location of the pointer identifies a currently 'selected node' of the UML diagram. In response to such input, the method dynamically highlights portions of the UML diagram depending on the location of the selected node. This highlighting process dynamically alters the appearances of nodes logically related to the selected node by highlighting nodes that are relatively closely related to the selected node differently (e.g., more brightly) than nodes that are not as relatively closely related to the selected node.

Growing community of inventors

Southampton, United Kingdom

Jonathan Limburn