OLAF: Towards Robust LLM-Based Annotation Framework in Empirical Software Engineering

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Imran, Mia Mohammad, Zaman, Tarannum Shaila
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911398741671936
author Imran, Mia Mohammad
Zaman, Tarannum Shaila
author_facet Imran, Mia Mohammad
Zaman, Tarannum Shaila
contents Large Language Models (LLMs) are increasingly used in empirical software engineering (ESE) to automate or assist annotation tasks such as labeling commits, issues, and qualitative artifacts. Yet the reliability and reproducibility of such annotations remain underexplored. Existing studies often lack standardized measures for reliability, calibration, and drift, and frequently omit essential configuration details. We argue that LLM-based annotation should be treated as a measurement process rather than a purely automated activity. In this position paper, we outline the \textbf{Operationalization for LLM-based Annotation Framework (OLAF)}, a conceptual framework that organizes key constructs: \textit{reliability, calibration, drift, consensus, aggregation}, and \textit{transparency}. The paper aims to motivate methodological discussion and future empirical work toward more transparent and reproducible LLM-based annotation in software engineering research.
format Preprint
id arxiv_https___arxiv_org_abs_2512_15979
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle OLAF: Towards Robust LLM-Based Annotation Framework in Empirical Software Engineering
Imran, Mia Mohammad
Zaman, Tarannum Shaila
Software Engineering
Artificial Intelligence
Large Language Models (LLMs) are increasingly used in empirical software engineering (ESE) to automate or assist annotation tasks such as labeling commits, issues, and qualitative artifacts. Yet the reliability and reproducibility of such annotations remain underexplored. Existing studies often lack standardized measures for reliability, calibration, and drift, and frequently omit essential configuration details. We argue that LLM-based annotation should be treated as a measurement process rather than a purely automated activity. In this position paper, we outline the \textbf{Operationalization for LLM-based Annotation Framework (OLAF)}, a conceptual framework that organizes key constructs: \textit{reliability, calibration, drift, consensus, aggregation}, and \textit{transparency}. The paper aims to motivate methodological discussion and future empirical work toward more transparent and reproducible LLM-based annotation in software engineering research.
title OLAF: Towards Robust LLM-Based Annotation Framework in Empirical Software Engineering
topic Software Engineering
Artificial Intelligence
url https://arxiv.org/abs/2512.15979