NAMeGEn: Creative Name Generation via A Novel Agent-based Multiple Personalized Goal Enhancement Framework

2025-11-19

AI Summary

Existing large language models struggle to simultaneously satisfy multiple personalized constraints—semantic meaning, phonological harmony, and cultural taboos—while providing interpretable aesthetic reasoning for Chinese baby naming, a canonical creative natural language generation (CNLG) task. Method: We propose the first multi-agent collaborative framework for CNLG, comprising three modules: (1) constraint parsing, (2) poetry-knowledge-enhanced iterative generation and evaluation, and (3) semantic consistency verification—enabling high-quality, zero-shot creative generation without fine-tuning. Contribution/Results: We introduce CBNames, the first benchmark for Chinese baby naming, along with culturally grounded evaluation metrics; integrate classical Chinese poetry corpora to enhance cultural awareness. Experiments demonstrate significant improvements over six baselines across creativity, compliance, and interpretability. To our knowledge, this is the first work to systematically achieve “generation-with-explanation” for short-text CNLG, delivering trustworthy, aesthetically justified outputs.

Technology Category

• Natural Language Processing: Generation
• Cognitive Modeling & Cognitive Systems: Computational Creativity
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Trained on diverse human-authored texts, Large Language Models (LLMs) unlocked the potential for Creative Natural Language Generation (CNLG), benefiting various applications like advertising and storytelling. Nevertheless, CNLG still remains difficult due to two main challenges. (1) Multi-objective flexibility: user requirements are often personalized, fine-grained, and pluralistic, which LLMs struggle to satisfy simultaneously; (2) Interpretive complexity: beyond generation, creativity also involves understanding and interpreting implicit meaning to enhance users'perception. These challenges significantly limit current methods, especially in short-form text generation, in generating creative and insightful content. To address this, we focus on Chinese baby naming, a representative short-form CNLG task requiring adherence to explicit user constraints (e.g., length, semantics, anthroponymy) while offering meaningful aesthetic explanations. We propose NAMeGEn, a novel multi-agent optimization framework that iteratively alternates between objective extraction, name generation, and evaluation to meet diverse requirements and generate accurate explanations. To support this task, we further construct a classical Chinese poetry corpus with 17k+ poems to enhance aesthetics, and introduce CBNames, a new benchmark with tailored metrics. Extensive experiments demonstrate that NAMeGEn effectively generates creative names that meet diverse, personalized requirements while providing meaningful explanations, outperforming six baseline methods spanning various LLM backbones without any training.

Problem

• Addressing multi-objective flexibility in creative language generation for personalized user requirements
• Overcoming interpretive complexity by generating meaningful explanations alongside creative content
• Enhancing short-form creative text generation for constrained tasks like Chinese baby naming

Innovation

• Agent-based framework optimizes multiple personalized objectives
• Iterative process extracts goals, generates names, evaluates results
• Enhances creativity using classical poetry corpus and new benchmark