Using String Metrics to Improve the Design of Virtual Conversational Characters: Behavior Simulator Development Study
Bibliographic record
Abstract
BACKGROUND: An emergency waiting room is a place where conflicts often arise. Nervous relatives in a hostile, unknown environment force security and medical staff to be ready to deal with some awkward situations. Additionally, it has been said that the medical interview is the first diagnostic and therapeutic tool, involving both intellectual and emotional skills on the part of the doctor. At the same time, it seems that there is something mysterious about interviewing that cannot be formalized or taught. In this context, virtual conversational characters (VCCs) are progressively present in most e-learning environments. OBJECTIVE: In this study, we propose and develop a modular architecture for a VCC-based behavior simulator to be used as a tool for conflict avoidance training. Our behavior simulators are now being used in hospital environments, where training exercises must be easily designed and tested. METHODS: We define training exercises as labeled, directed graphs that help an instructor in the design of complex training situations. In order to increase the perception of talking to a real person, the simulator must deal with a huge number of sentences that a VCC must understand and react to. These sentences are grouped into sets identified with a common label. Labels are then used to trigger changes in the active node of the graph that encodes the current state of the training exercise. As a consequence, we need to be able to map every sentence said by the human user into the set it belongs to, in a fast and robust way. In this work, we discuss two different existing string metrics, and compare them to one that we use to assess a designed exercise. RESULTS: Based on the similarities found between different sets, the proposed metric provided valuable information about ill-defined exercises. We also described the environment in which our programs are being used and illustrated it with an example. CONCLUSIONS: Initially designed as a tool for training emergency room staff, our software could be of use in many other areas within the same environment. We are currently exploring the possibility of using it in speech therapy situations.
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 | 0.006 | 0.041 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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 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".