Great Expectations and EPIC Fails: A Computational Perspective on Irony and Sarcasm
Bibliographic record
Abstract
We typically believe that irony is a completely human affair, but there have been interesting attempts to create computational models of irony use and understanding. This chapter presents an overview of some of these models, especially as implemented as conversational agents. One of the beauties, and major challenges, of computer modeling is that it forces researchers to make concrete decisions on how best to implement some linguistic observation or theoretical idea (e.g., how to create a workable model of echoic mention, pretense, or what is meant by incongruity). Veale presents his EPIC model in which an expectation (E) predicts a property (P) of an instance (I) of concept (C) that can get upended by an ironic utterance. This model provides a quantifiable view of what it means for an ironic utterance to achieve its desired effect on an audience. The success of an ironic utterance hinges on its capacity to highlight the failure of a reasonable expectation. The effectiveness of this computational model was partly assessed by obtaining human judgments about the meaning and quality of different ironic utterances in varying contexts that are suggestive of different expectations. In this way, Veale’s work offers insights as to how engineering solutions may be very informative about the way irony functions in human communication.
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 | 0.002 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.002 | 0.008 |
| Scholarly communication | 0.004 | 0.007 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.006 | 0.001 |
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".