NLBSE 2026

Accepted Papers

• Automating Metadata Extraction from Software Engineering Contracts Using a RAG Framework
  Momojit Biswas, Chirag Jain and Preethu Rose Anish
• From Issues to Insights: RAG-based Explanation Generation from Software Engineering Artifacts
  Daniel Pöttgen, Mersedeh Sadeghi, Max Unterbusch and Andreas Vogelsang
• Graph-RAG for Contributor Discovery on GitHub
  Usama Nisar and Malaika Farooq
• Bridging Communication Gaps: Inclusive Conversational Agents for Requirement Engineering
  Munazza Zaib
• Bridging Correctness and Quality in LLM Code Generation: A Hybrid Generate-and-Refine Pipeline
  Nouman Munib
• How should self-deprecation comments be classified? A toxicity analysis study on Zephyr
  Satyanarayana Chowdary Kadiyala, Jaydeb Sarker and Bianca Trinkenreich

Workshop program

NLBSE'26 will be co-located with ICSE'26 in Rio de Janeiro, Brazil.

Room: Oceania VI (shared with NSE 2026 — 2nd International Workshop on Neuro-Symbolic Software Engineering)

NSE occupies the morning session (09:00–12:00). NLBSE starts before lunch (12:00–12:30) and continues through the afternoon (14:00–18:30).

Morning — NSE Sessions

Time	Activity
09:00–09:10	NSE Opening
09:10–10:10	NSE Keynote + Panel Discussion
10:10–10:30	NSE Paper Presentation (15+5 min)
Coffee break / Community pictures (NSE/NLBSE)
11:00–12:00	NSE Paper Presentations — 3 papers (15+5 min each) + NSE Closing

Pre-Lunch — NLBSE Opening

Time	Activity
12:00–12:05	NLBSE Opening
12:05–12:30	Keynote: "Answering Questions About Open Source Projects" — Marco A. Gerosa (Northern Arizona University, USA)

Lunch break (12:30–14:00)

Afternoon — NLBSE Sessions

Time	Activity
14:00–14:30	Keynote: "Beyond Off-the-Shelf: Building Reliable NLP Tools for Software Engineering" — Nicole Novielli (University of Bari, Italy)
14:30–15:00	Long Papers (12+3 min each) • Automating Metadata Extraction from SE Contracts Using a RAG Framework — M. Biswas, C. Jain, P. R. Anish • From Issues to Insights: RAG-based Explanation Generation from SE Artifacts — D. Pöttgen, M. Sadeghi, M. Unterbusch, A. Vogelsang
15:00–15:20	Short Papers (8+2 min each) • Bridging Communication Gaps: Inclusive Conversational Agents for Requirement Engineering — M. Zaib • How should self-deprecation comments be classified? A toxicity analysis study on Zephyr — S. C. Kadiyala, J. Sarker, B. Trinkenreich
15:20–15:30	NLBSE Awards and Message from the Chairs
Coffee break
16:00–16:05	NLBSE Tool Competition Opening
16:05–16:33	Tool Competition: Code Comment Classification (5+2 min each) • High-quality data augmentation for code comment classification — T. Borsani, A. Rosani, G. Di Fatta • TraCC: Efficient Multi-Label Code Comment Classification Through Knowledge Distillation and Adaptive Thresholding — P. S. U. Shah, A. Ashfaq, A. Fasseh, D. Shah • X-LoRA MME: Multi-Model Ensemble with Mixture of Experts for Code Comment Classification — M. A. Haider, A. Bulbul, A. Ahmed, N. F. Shahid, M. I. Abedin • Distilling Semantics: Efficient Multi-label Code Comment Classification — M. A. Majeed, A. B. Asim
16:33–16:38	Tool Competition Awards and Closing
16:38–18:28	Tutorial (IN-PERSON ONLY): "Building an LLM-based coding copilot" — Pedro Mário Cruz e Silva (NVIDIA)
18:28–18:30	NLBSE Closing