GCOF: Self-iterative Text Generation for Copywriting Using Large Language Model

arXiv:2402.13667v1 Announce Type: new
Abstract: Large language models(LLM) such as ChatGPT have substantially simplified the generation of marketing copy, yet producing content satisfying domain specific requirements, such as effectively engaging customers, remains a significant challenge. In this work, we introduce the Genetic Copy Optimization Framework (GCOF) designed to enhance both efficiency and engagememnt of marketing copy creation. We conduct explicit feature engineering within the prompts of LLM. Additionally, we modify the crossover operator in Genetic Algorithm (GA), integrating it into …

abstract arxiv challenge chatgpt copy copywriting cs.cl customers domain framework iterative language language model language models large language large language model large language models llm marketing optimization requirements simplified text text generation type work