Probabilistic relaxed unification formalism and its application in question answering
Bibliographic record
Abstract
We explore the problem of graph unification of mismatched graphs. One of the assumptions of classical unification is that the knowledge base is complete and accurate, which seldom applies to the real world. We present relaxed unification as an alternative approach where the assumptions of classical unification are relaxed. Relaxed unification replaces the binary success or failure outcome of classical unification with a real number quantifying the correctness of the result. We provide a theoretical framework for relaxed unification by defining the relaxed unification formalism and present an algorithm for relaxed unification. We extend the formalism to probabilistic relaxed unification and devise an evaluation function that assigns correctness value to the result of the unification based on random walks in finite Markov chains. We present a modular framework for question answering and realize it in the implementation of a Relaxed Unification Question Answering system prototype. The system relies on the Jellyfish question answering system for question analysis and extraction of candidate answer. Retrieval of relevant documents is handled by the Apache Lucene retrieval engine. A semantic representation of the question and the candidate answers is produced using DELPH-IN tools and resource. Finally, the Relaxed Unification Module unifies the semantic representation of the candidate answers with that of the question, evaluates the correctness of the results, and ranks the final answers accordingly. Our approach is empirically validated through a series of cases drawn from real world questions and data collection. The validation cases substantiate that our system provides satisfactory results on the chosen dataset within the system limitations. They provide a detailed walkthrough of the system operation, demonstrate the granularity of the correctness function, present a method for incorporating a word similarity measure in the computation of the correctness function, and demonstrate the system limitations.
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 | 0.017 | 0.050 |
| Meta-epidemiology (narrow) | 0.001 | 0.002 |
| Meta-epidemiology (broad) | 0.002 | 0.005 |
| Bibliometrics | 0.006 | 0.006 |
| Science and technology studies | 0.002 | 0.006 |
| Scholarly communication | 0.005 | 0.013 |
| Open science | 0.005 | 0.007 |
| Research integrity | 0.003 | 0.006 |
| Insufficient payload (model declined to judge) | 0.005 | 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".