TransitTalk: Large language model-based digital assistants for enhancing transit customer experience and staff performance
Bibliographic record
Abstract
Large-language models (LLMs) have transformed natural-language understanding, generation, and reasoning, yet their potential within public-transit operations remains largely untapped. This study introduces TransitTalk, a retrieval-augmented framework that positions an LLM as an intelligent intermediary between natural-language requests and the heterogeneous databases that underpin modern transit networks. The framework first interprets rider or staff intent, then issues structured queries to relevant data sources, and finally produces context-aware conversational responses. Chain-of-thought prompting and vector-database retrieval enable the LLM to operate simultaneously as a dialogue agent, domain expert, and bespoke formatter, thereby bridging the gap between technical data schemas and user-friendly information delivery. The paper demonstrates the practicality of the framework through three fully implemented prototypes. Tweet Writer automatically drafts clear, multi-platform service-alert messages; Trip Advisor generates personalized, constraint-aware itineraries; and Policy Navigator delivers concise explanations of fares, accessibility provisions, and service regulations. Collectively, these applications illustrate how an LLM layer can streamline communications, reduce staff workload, and enhance the passenger information experience without disrupting existing data pipelines. All source code are openly available.
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 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.004 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.012 | 0.003 |
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".