Browse Prior Art Database

Using handlers in a topic space to establish non-hierarchic relationships

IP.com Disclosure Number: IPCOM000125712D
Original Publication Date: 2005-Jun-14
Included in the Prior Art Database: 2005-Jun-14
Document File: 2 page(s) / 39K

Publishing Venue

IBM

Abstract

Topic spaces are used within publish/subscribe messaging systems. They define a hierarchic tree structure of topics to be used by applications when subscribing to sets of messages. Conceptually, the relationships between messages within a topic space may be considered "vertical" since they are constrained topologically by the tree structure. In complex information spaces, the simplistic hierarchic tree structure for topics is inadequate and relationships between messages become multi-dimensional. This disclosure proposes a mechanism of software handlers that can be used to generate more complex "non-vertical" relationships between messages.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 53% of the total text.

Page 1 of 2

Using handlers in a topic space to establish non -hierarchic relationships

Disclosed is a mechanism for creating multi-dimensional relationships between messages in a publish/subscribe messaging system using software handlers.

  Topic spaces are used within publish/subscribe messaging systems. They define a hierarchic set of related topics that can be individually attached to messages when published. Application clients may subscribe to sets of topics. A broker receives a published message and routes it to the client(s) that subscribed to the topic. Topic spaces are hierarchic and have a tree structure. There may be many different topic spaces within the messaging infrastructure, but they all have a common root. Conceptually, the relationships between messages within a topic space may be considered as "vertical" since they are constrained topologically by a hierarchic tree structure.

  In complex information spaces, the simplistic hierarchic tree structure is inadequate. Relationships become multi-dimensional, creating a network of relationships between individual "nuggets" of information. Consider a topic space for the analysis of musical events. There may be several different topic space trees (though with a common root);

  A simple Midi topic space for primitive musical events - whether a note was depressed or not.:

  /Midi/Short/NoteOn /Midi/Short/NoteOff
An harmonic topic space used to publish messages resulting from a harmonic analysis of the musical stream of events:

  /Harmonic/Chord/Maj /Harmonic/Chord/Min /Harmonic/Interval/Perfect /Harmonic/Interval/Dim
The analysis will take as its input a stream of Midi note events and publish its results. The published analysis results (inthe /Harmonic/ topic space) will be linked to the note events (in the /Midi/ topic space). Therefore it would be advantageous to create links in the information space between the analysis results and the events that triggered it. This requires links to be created between leaf elements or sub-trees in one topic space to a set of elemen...