Using Transformer Models and Textual Analysis for Log Parsing
Bibliographic record
Abstract
Log parsing has become an essential tool for extracting valuable information from a vast volume of log lines. It involves identifying standard patterns and extracting templates from these lines, enabling researchers and companies to employ advanced mining techniques like log deduplication and log anomaly detection. However, existing log parsing approaches have limitations. They often operate on small batches of log text and lack consideration for the entire context, necessitating prior knowledge of the log dataset. Furthermore, there is a scarcity of practical experience reports on the utilization of these log parsing approaches in the literature. In our paper, we address these challenges by proposing a novel log parsing approach that combines Transformers with a customized textual analysis. This textual analysis balances the density clustering of similar log lines, the calculation of word frequencies inside each cluster and the presence of the words inside an English vocabulary to parse new log lines. Our method outperforms existing parsing methods in terms of accuracy and operates as an unsupervised learning approach, eliminating the need for prior knowledge. Additionally, we present a proof-of-concept application of our parsing method in an industrial setting, showcasing its practical implementation.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.004 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.004 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".