Mutation Testing of Event Processing Queries
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Event processing queries are intended to process continuous event streams. These queries are partially similar to traditional SQL queries, but provide the facilities to express rich features (e.g., pattern expression, sliding window of length and time). An error while implementing a query may result in abnormal program behaviors and lost business opportunities. Moreover, queries can be generated with unsanitized inputs and the structure of intended queries might be altered. Thus, a tester needs to test the behavior of queries in presence of malicious inputs. Mutation testing has been found to be effective to assess test suites quality and generating new test cases. Unfortunately, there is no effort to perform mutation testing of event processing queries. In this work, we propose mutation-based testing of event processing queries. We choose Event Processing Language (EPL) as our case study and develop necessary mutation operators and killing criteria to generate high quality event streams and malicious inputs. Our proposed operators modify different features of EPL queries (pattern expression, windows of length and time, batch processing of events). We develop an architecture to generate mutants for EPL and perform mutation analysis. We evaluate our proposed EPL mutation testing approach with a set of developed benchmark containing diverse types EPL queries. The evaluation results indicate that the proposed operators and mutant killing criteria are effective to generate test cases capable of revealing anomalous program behaviors (e.g., event notification failure, delay of event reporting, unexpected event), and SQL injection attacks. Moreover, the approach incurs less manual effort and can complement other testing approach such as random testing.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 it