---
_id: '63524'
abstract:
- lang: eng
  text: Recommendation systems are essential for delivering personalized content across
    e-commerce and streaming services. However, traditional methods often fail in
    cold-start scenarios where new items lack prior interactions. Recent advances
    in large language models (LLMs) offer a promising alternative. In this paper,
    we adopt the retrieve-and-recommend framework and propose to fine-tune the LLM
    jointly on warm-and cold-start next-item recommendation tasks, thus, mitigating
    the need for separate models for both item types. We computationally compare zero-shot
    prompting, in-context learning, and fine-tuning using the same LLM backbone, and
    benchmark them against strong PLM-based baselines. Our findings provide practical
    insights into the trade-offs between accuracy and computational cost of these
    methods for next-item recommendation. To enhance reproducibility, we release the
    source code under https://github. com/HayaHalimeh/LLMs-For-Next-Item-Recommendation.git.
author:
- first_name: Haya
  full_name: Halimeh, Haya
  id: '87673'
  last_name: Halimeh
- first_name: Florian
  full_name: Freese, Florian
  last_name: Freese
- first_name: Oliver
  full_name: Müller, Oliver
  last_name: Müller
citation:
  ama: 'Halimeh H, Freese F, Müller O. LLMs For Warm and Cold Next-Item Recommendation:
    A Comparative Study across Zero-Shot Prompting, In-Context Learning and Fine-Tuning.
    In: <i>International Conference on Information Systems Development</i>. University
    of Gdansk, Department of Business Informatics &#38; University of Belgrade, Faculty
    of Organizational Sciences; 2025. doi:<a href="https://doi.org/10.62036/isd.2025.68">10.62036/isd.2025.68</a>'
  apa: 'Halimeh, H., Freese, F., &#38; Müller, O. (2025). LLMs For Warm and Cold Next-Item
    Recommendation: A Comparative Study across Zero-Shot Prompting, In-Context Learning
    and Fine-Tuning. <i>International Conference on Information Systems Development</i>.
    <a href="https://doi.org/10.62036/isd.2025.68">https://doi.org/10.62036/isd.2025.68</a>'
  bibtex: '@inproceedings{Halimeh_Freese_Müller_2025, title={LLMs For Warm and Cold
    Next-Item Recommendation: A Comparative Study across Zero-Shot Prompting, In-Context
    Learning and Fine-Tuning}, DOI={<a href="https://doi.org/10.62036/isd.2025.68">10.62036/isd.2025.68</a>},
    booktitle={International Conference on Information Systems Development}, publisher={University
    of Gdansk, Department of Business Informatics &#38; University of Belgrade, Faculty
    of Organizational Sciences}, author={Halimeh, Haya and Freese, Florian and Müller,
    Oliver}, year={2025} }'
  chicago: 'Halimeh, Haya, Florian Freese, and Oliver Müller. “LLMs For Warm and Cold
    Next-Item Recommendation: A Comparative Study across Zero-Shot Prompting, In-Context
    Learning and Fine-Tuning.” In <i>International Conference on Information Systems
    Development</i>. University of Gdansk, Department of Business Informatics &#38;
    University of Belgrade, Faculty of Organizational Sciences, 2025. <a href="https://doi.org/10.62036/isd.2025.68">https://doi.org/10.62036/isd.2025.68</a>.'
  ieee: 'H. Halimeh, F. Freese, and O. Müller, “LLMs For Warm and Cold Next-Item Recommendation:
    A Comparative Study across Zero-Shot Prompting, In-Context Learning and Fine-Tuning,”
    2025, doi: <a href="https://doi.org/10.62036/isd.2025.68">10.62036/isd.2025.68</a>.'
  mla: 'Halimeh, Haya, et al. “LLMs For Warm and Cold Next-Item Recommendation: A
    Comparative Study across Zero-Shot Prompting, In-Context Learning and Fine-Tuning.”
    <i>International Conference on Information Systems Development</i>, University
    of Gdansk, Department of Business Informatics &#38; University of Belgrade, Faculty
    of Organizational Sciences, 2025, doi:<a href="https://doi.org/10.62036/isd.2025.68">10.62036/isd.2025.68</a>.'
  short: 'H. Halimeh, F. Freese, O. Müller, in: International Conference on Information
    Systems Development, University of Gdansk, Department of Business Informatics
    &#38; University of Belgrade, Faculty of Organizational Sciences, 2025.'
date_created: 2026-01-07T13:36:53Z
date_updated: 2026-01-07T13:47:43Z
department:
- _id: '195'
- _id: '196'
doi: 10.62036/isd.2025.68
language:
- iso: eng
main_file_link:
- open_access: '1'
oa: '1'
publication: International Conference on Information Systems Development
publication_identifier:
  issn:
  - 2938-5202
publication_status: published
publisher: University of Gdansk, Department of Business Informatics & University of
  Belgrade, Faculty of Organizational Sciences
status: public
title: 'LLMs For Warm and Cold Next-Item Recommendation: A Comparative Study across
  Zero-Shot Prompting, In-Context Learning and Fine-Tuning'
type: conference
user_id: '87673'
year: '2025'
...