It resource event situation classification and semantics

• US 7,461,044 B2
• Filed: 04/27/2005
• Issued: 12/02/2008
• Est. Priority Date: 04/27/2005
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Minerley, Kevin Gordon, Salahshour, Abdolreza, Sheng, Ma, Loewenstern, David Matthew
• Primary Examiner(s): Vincent; David
• Assistant Examiner(s): Brown, Jr.; Nathan H
• Application Number:
  US11/117,151
• Publication Number:
  US 20070005535A1
• Time in Patent Office: 1,315 Days
• Field of Search: 706/47, 382/224
• US Class Current: 706/47
• CPC Class Codes: G06F 9/542 Event management; Broadcasting; Multicasting; Notifications

Abstract

Classifying a message includes receiving a message to be classified, wherein the message includes a message identifier. If the message identifier uniquely maps to a corresponding classification category, the message is labeled with the identified classification category. If the message identifier does not map directly to a corresponding classification category, the message to be classified is parsed and a plurality of features from the parsed message are identified, wherein at least one classification rule is compared to the plurality of features. Each classification rule that matches to the plurality of features is rated and a classification category is identified from the rating, wherein the message is labeled with the identified classification category.

Claims

1. A computer implemented method of classifying a message, the method comprising:

receiving a message to be classified, said message having a message identifier;

determining if said message identifier uniquely maps to a corresponding classification category;

labeling the message with the identified classification category if said message identifier maps directly to a corresponding classification category;

parsing the message to be classified and identifying a plurality of features from the parsed message if said message identifier does not map directly to a corresponding classification category;

comparing at least one classification rule to the plurality of features if said message identifier does not map directly to a corresponding classification category;

rating each classification rule that matches to said plurality of features;

identifying a classification category from said rating; and

labeling the message with the identified classification category.

2. The computer implemented method of claim 1, the method further comprising:

comparing a second classification rule to the plurality of features.

3. The computer implemented method of claim 1, the method further comprising:

receiving a plurality of messages;

parsing the plurality of messages;

identifying features from the plurality of messages;

selecting a common classification category semantic;

matching at least some of the identified features with the common classification category semantic; and

generating a classification rule from the identified features matched with the common classification category semantic.

4. The computer implemented method of claim 1, the method further comprising:

receiving a plurality of messages;

parsing the plurality of messages;

identifying features from the plurality of messages;

selecting a plurality of common classification category semantics;

matching at least some of the identified features with the respective plurality of common classification category semantics;

identifying messages corresponding to the at least some of the identified features; and

generating a classification category to message map.

5. The computer implemented method of claim 3, the method further comprising:

selecting a plurality common classification category semantics;

identifying at least some of the plurality of messages having features matching the plurality of common classification category semantics;

dividing the identifying at least some of the plurality of messages into respective clusters of messages each having a unique common classification category semantic;

reviewing features from messages in the respective clusters; and

revising the classification rule.