Abstract:
This research aims to analyze the in-demand skills within Thailands software
industry, focusing on both technical skills and soft skills as found in real job postings.
Natural Language Processing (NLP) techniques were applied to extract insights from job descriptions. Data were collected from JobThai and LinkedIn, totaling records during 20232024. The dataset underwent cleaning, tokenization, deduplication, and frequency analysis to identify dominant skills across different job roles. The findings revealed that programming languages such as Java, SQL, and Python were the most frequently required technical skills across nearly all positions. On the soft skill side, communication, collaboration, and problem-solving consistently ranked as the top three skills in every job category.
Furthermore, the study compared the performance of various machine learning
models for job role prediction, including Random Forest, Na?ve Bayes, and XGBoost, as well as a deep learning model (LSTM). Among these, Random Forest achieved the highest performance. The results of this study can be used to inform curriculum development aligned with labor market needs and support the use of predictive models to recommend future skill development strategies.