Study Introduces LLM‑TOPSIS Framework for Automated Software Engineer Hiring

Global: Study Introduces LLM‑TOPSIS Framework for Automated Software Engineer Hiring Researchers have unveiled an automated personnel selection system that integrates large language models with a fuzzy TOPSIS decision‑making approach to rank software engineering applicants, reporting accuracy of up to 91% for both the Experience and Overall evaluation attributes.

Dataset Construction

The authors compiled a distinctive dataset by aggregating publicly available LinkedIn profiles, capturing education, work experience, skill sets, and self‑descriptions. Expert assessments were added to serve as reference standards for each candidate.

Hybrid LLM‑TOPSIS Method

The proposed LLM‑TOPSIS framework combines a fine‑tuned DistilRoBERTa model with a fuzzy TOPSIS algorithm enhanced by triangular fuzzy numbers, which model the inherent ambiguity in criteria weights and scores.

Performance Results

Evaluation against human expert rankings showed close alignment, with the system achieving 91% accuracy on the Experience attribute and the Overall attribute, indicating strong concordance with professional judgments.

Practical Implications

According to the authors, the approach promises greater scalability and consistency in recruitment while aiming to mitigate subjective bias, potentially transforming conventional hiring workflows.

Future Directions

Ongoing work will focus on expanding the dataset, improving model interpretability, and validating the system in real‑world recruitment settings to assess practical applicability. This report is based on information from arXiv, licensed under Academic Preprint / Open Access. Based on the abstract of the research paper. Full text available via ArXiv.