Log Stream Fingerprinting via Matrix-Based Feature Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining device types from log streams are resource-intensive, especially as the number of log streams increases, and require significant resources for parsing, which is inefficient.
Innovation Solution
A method that identifies features in log streams from unknown devices by comparing them to known device types using a matrix-based approach, where features are tokenized, weighted, and matched against a fingerprint map matrix to determine probabilities of device types, reducing the need for extensive resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If regular expression-based pattern matching is used to detect patterns in data streams, then pattern detection capability is improved, but resource consumption increases significantly
Solution Approach 1:
The patent segments the log stream analysis into distinct phases: tokenization of log entries into discrete elements, extraction of specific features from tokens, and comparison against a fingerprint map. This segmentation replaces holistic regular expression matching with targeted feature-based comparison, reducing resource consumption while maintaining detection capability.
Solution Approach 2:
The patent extracts only the relevant features from log streams that are characteristic of specific device types, rather than processing entire log entries. By taking out and comparing only these distinguishing features against a fingerprint map, the system achieves efficient pattern detection without the overhead of comprehensive regular expression-based analysis.
2Measurement precision
If extensive parsing is performed on log streams to determine device types, then device type identification accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary action by pre-computing and storing fingerprint maps for known device types during an offline phase. These fingerprint maps contain pre-extracted features and characteristics that can be quickly compared against incoming log streams. This preliminary preparation eliminates the need for extensive real-time parsing, significantly reducing processing time while maintaining identification accuracy.
Solution Approach 2:
The patent creates simplified copies of device type characteristics in the form of fingerprint maps. Instead of parsing and analyzing entire log streams in real-time, the system compares incoming logs against these pre-created fingerprint copies, achieving fast and accurate device type identification without extensive processing.
3Use of energy by moving object
If feature-based matching with probability determination is implemented, then resource usage is reduced, but implementation complexity increases
Solution Approach 1:
The patent changes the parameters of log analysis from holistic text processing to structured feature extraction with numerical probability scoring. By transforming log data into discrete tokens, extracting specific features, and computing probability scores based on feature matching, the system achieves resource-efficient processing. The complexity is managed through systematic parameter transformation rather than ad-hoc analysis.
Data Source
AI summary
The method includes identifying a log stream from a first device, wherein the first device is of an unknown device type. The method further includes identifying features in the log stream from the first device. The method further includes determining a probability that the log stream from the first device was created by a specific device type based on matching of features of the identified log stream from the first device and features in log streams from a plurality of known device types.