| dc.description.abstract | Studies on the international construction market have been limited to expanding the scope of academics and practices because of data accessibility and timeliness. With the recent advancement of natural language processing (NLP) technologies, it becomes possible to extract on-time information from online news articles automatically. As a point of departure for developing a text-based information extraction model, this study aims to develop a named entity recognition (NER) model that automatically detects active players from news articles in the international construction industry. NER is an essential subtask of information extraction that automatically identifies key elements and classifies them into predefined categories. The proposed model detects owners, contractors, and consultants from news articles. The performance of the experiment was measured by a micro average F1 score of 85.8% with precision and recall values of 84.2% and 87.4%, respectively. This study contributes to investigating international market participants in a timely way with enhanced data accessibility. Therefore, the following studies will enlarge the NER approach to recognize “Who contacts whom,” “Who claims whom,” and “What delays what projects,” which will lead to extracting more valuable information automatically in the future. | |
