# From User Preferences to Optimization Constraints Using Large Language Models

arXiv: https://arxiv.org/abs/2503.21360 · HTML: https://mlsanity.com/paper/2503.21360/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 9 of its 22 references that are in the map.
- First posted on arXiv 27 Mar 2025.

## Builds on (most cited first, 9 of 9)

- [NL4Opt Competition: Formulating Optimization Problems Based on Their Natural Language Descriptions](https://mlsanity.com/paper/2303.08233/) (2023)
- [Rasa: Open Source Language Understanding and Dialogue Management](https://mlsanity.com/paper/1712.05181/) (2017)
- [Out of the BLEU: how should we assess quality of the Code Generation models?](https://mlsanity.com/paper/2208.03133/) (2022)
- [Augmenting Operations Research with Auto-Formulation of Optimization Models from Problem Descriptions](https://mlsanity.com/paper/2209.15565/) (2022)
- [DIET: Lightweight Language Understanding for Dialogue Systems](https://mlsanity.com/paper/2004.09936/) (2020)
- [LLaMAntino: LLaMA 2 Models for Effective Text Generation in Italian Language](https://mlsanity.com/paper/2312.09993/) (2023)
- [Synthesizing mixed-integer linear programming models from natural language descriptions](https://mlsanity.com/paper/2311.15271/) (2023)
- [Advanced Natural-based interaction for the ITAlian language: LLaMAntino-3-ANITA](https://mlsanity.com/paper/2405.07101/) (2024)
- [Highlighting Named Entities in Input for Auto-Formulation of Optimization Problems](https://mlsanity.com/paper/2212.13201/) (2022)

## Abstract (arXiv)

This work explores using Large Language Models (LLMs) to translate user preferences into energy optimization constraints for home appliances. We describe a task where natural language user utterances are converted into formal constraints for smart appliances, within the broader context of a renewable energy community (REC) and in the Italian scenario. We evaluate the effectiveness of various LLMs currently available for Italian in translating these preferences resorting to classical zero-shot, one-shot, and few-shot learning settings, using a pilot dataset of Italian user requests paired with corresponding formal constraint representation. Our contributions include establishing a baseline performance for this task, publicly releasing the dataset and code for further research, and providing insights on observed best practices and limitations of LLMs in this particular domain