Log Tracing via Temporal Association Degree Computation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for tracing log data require learning in advance or the use of a common identifier, making it difficult to introduce them into existing systems.

Innovation Solution

A log tracing apparatus that computes an association degree between log data types based on their occurrence times, allowing for association without prior learning or a common identifier, using components like an acquisition unit, association degree computation unit, and log association unit.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If learning in advance is performed to identify association between log data, then measurement precision of log association is improved, but device complexity and ease of operation deteriorate due to requiring test data and pre-processing

Engineering Contradiction:

Improvelog association identification accuracyVSAvoidsystem introduction difficulty

Core Design Contradiction:

Measurement precisionVSEase of operation

Solution Approach 1:

The system automatically computes association degrees between log data types using occurrence time patterns without requiring manual pre-learning or test data preparation. The association degree computation unit self-service computes probabilities and identifies correlations autonomously when log data is input, eliminating the need for operators to perform advance learning procedures.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system pre-computes and stores occurrence time patterns and association degrees between different log data types during system operation. These pre-computed association degrees are then reused for rapid log association identification without requiring repeated learning processes, enabling both accuracy and ease of operation.

Inventive Principle:

Principle #10Preliminary action

2Measurement precision

If a common identifier is used to identify association between log data, then measurement precision is improved, but adaptability deteriorates due to requiring standardized identifiers across systems

Engineering Contradiction:

Improvelog association identification accuracyVSAvoidexisting system compatibility

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

Instead of relying on common identifiers, the system changes the identification parameter to use occurrence time patterns. The association degree computation unit computes probabilities based on temporal relationships between log events, allowing the system to identify associations without requiring standardized identifier formats across different log sources.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The system introduces occurrence time patterns as an intermediary element that mediates between different log data types. By computing association degrees based on temporal patterns rather than direct identifier matching, the system enables log association across heterogeneous systems that may use different identifier schemes.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS9900331B2Log tracing apparatus and non-transitory computer-readable medium storing a log tracing program

Publication Date: 2018.02.20 MITSUBISHI ELECTRIC CORP

AI summary

An association degree computation unit computes a same period probability between respective types of events that occurred during a same period, based on occurrence times of log data, and computes an association degree between the respective types, using the same period probability. A corresponding type identification unit identifies the types corresponding to each other, using the association degree computed by the association degree computation unit. The log association unit generates synthesized log data that makes association between the log data of the corresponding types identified by the corresponding event identification unit and with the occurrence times thereof included in the same period.