Representing argumentation schemes with Constraint Handling Rules (CHR)
Bibliographic record
Abstract
We present a high-level declarative programming language for representing argumentation schemes, where schemes represented in this language can be easily validated by domain experts, including developers of argumentation schemes in informal logic and philosophy, and serve as executable specifications for automatically constructing arguments, when applied to a set of assumptions. Since argumentation schemes are defeasible inference rules, both premises and conclusions of schemes can be second-order schema variables, i.e. without a fixed predicate symbol. Thus, while particular schemes can be and have been implemented in computer programs, in general argumentation schemes cannot be represented as executable specifications using logic programming languages based on first-order logic, such as Prolog. Moreover, even if the conclusion (head) of Prolog rules could be second-order variables, a depth-first, backward-chaining search strategy, as typically used in logic programming, would usually cause such programs to not terminate, since every goal would match the head of such a scheme, including all goals created by instantiating the body of the same scheme. The language for representing argumentation schemes presented here, for the purpose of automatically constructing arguments, uses Constraint Handling Rules (CHR), a declarative, Turing complete, forwards-chaining, rule-based programming language introduced by Thom Frühwirth in 1991. CHR is attractive for representing and implementing argumentation for several reasons, including: 1) Inference rules, rewrite rules, sequents, proof rules, and logical axioms can be directly written in CHR. 2) The execution of CHR rules can be interrupted and restarted at any time, with intermediate results approximating the final solution, and 3) Constraints can be input incrementally as they become known, during rule execution, without requiring recomputation. These three properties of CHR appear attractive for representing and implementing argumentation schemes. Since argumentation schemes are (defeasible) inference rules, the ability of CHR to represent inference rules directly would appear to be quite useful. The ability to stop the computation and produce approximate results is compatible with one objective of argumentation, to provide a principled method for supporting approximate reasoning with limited resources. Because argumentation typically takes place in dialogs, with evidence and arguments brought forward and asserted by the participants incrementally, during the course of the dialog, CHR’s ability to handle new information, incrementally introduced during the computation, may be useful. This new rule language for representing argumentation schemes is validated by using it to represent twenty representative argumentation schemes.
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 | 0.004 | 0.011 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.003 |
| Bibliometrics | 0.003 | 0.003 |
| Science and technology studies | 0.001 | 0.003 |
| Scholarly communication | 0.005 | 0.007 |
| Open science | 0.004 | 0.003 |
| Research integrity | 0.003 | 0.003 |
| Insufficient payload (model declined to judge) | 0.007 | 0.002 |
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".