This repository documents the findings of a pioneering thesis research project focused on investigating the feasibility of utilizing Large Language Models (LLMs) to generate advertising sentences from keywords.
Objective: To explore the potential of LLMs in revolutionizing advertising content creation.
Methodology: The study involved conducting experiments with a sample size of 100 individuals to evaluate the effectiveness of sentences generated by the LLM compared to human-generated ones.
Key Finding: Contrary to expectations, the study revealed that sentences produced by the LLM were preferred over original human-generated sentences.
These findings suggest a significant shift in the landscape of advertising content creation, indicating the potential for LLMs to revolutionize the industry.
Repository Contents:
Poster: Detailed insights into the research methodology, results, and implications.
Code: Implementation of the LLM model used in the research, along with instructions for replication and further experimentation.

Jupyter Notebook
91.0%
Python
5.6%
Shell
3.4%
This repository documents the findings of a pioneering thesis research project focused on investigating the feasibility of utilizing Large Language Models (LLMs) to generate advertising sentences from keywords.
Objective: To explore the potential of LLMs in revolutionizing advertising content creation.
Methodology: The study involved conducting experiments with a sample size of 100 individuals to evaluate the effectiveness of sentences generated by the LLM compared to human-generated ones.
Key Finding: Contrary to expectations, the study revealed that sentences produced by the LLM were preferred over original human-generated sentences.
These findings suggest a significant shift in the landscape of advertising content creation, indicating the potential for LLMs to revolutionize the industry.
Repository Contents:
Poster: Detailed insights into the research methodology, results, and implications.
Code: Implementation of the LLM model used in the research, along with instructions for replication and further experimentation.

Jupyter Notebook
91.0%
Python
5.6%
Shell
3.4%