Experimental Analysis of Large Language Models in Crime Classification and Prediction
Bibliographic record
Abstract
Increasing crime rates and evolving challenges in law enforcement have raised the need for innovative solutions, leading to the emergence of smart policing. This paradigm shift incorporates artificial intelligence (AI) with a specific focus on machine learning (ML) as a pivotal tool for data analysis, pattern recognition, and proactive crime forecasting. Large-language models (LLMs) as a subset of generative AI have been used in different domains, such as financial, medical, legal, and agricultural applications. However, the abilities and possibilities of adopting LLMs for smart policing applications such as crime classification remain unexplored. This paper explores the transformative potential of BART, GPT-3, and GPT-4, three state-of-the-art LLMs, in the domain of crime analysis and predictive policing. Utilizing diverse methods such as zero-shot prompting, few-shot prompting, and fine-tuning, this paper evaluates the performance of these models on state-of-the-art datasets from two major cities: San Francisco and Los Angeles. The goal is to demonstrate the adaptability of LLMs and their capacity to revolutionize conventional crime analysis practices. The paper also provide a comparative analysis of the aforementioned methods on the GPT series model and BART, in addition to ML techniques, showing that GPT models are more suitable for crime classification in most of our experimental scenarios.
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 | 0.011 | 0.043 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.002 | 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".