A conversation-based system is proposed for supporting assessors in performing existing school building inspections. School building safety is a pressing issue; however, some difficulties in the overall process require solutions or improvements, including the complexity of building inspection tasks, the restrictions posed by the paperwork process, and the ineffectiveness of the management of existing school building inspections. In this study, we developed a conversation-based building inspection support system to reduce such problems, with the proposed system notifying and guiding assessors to complete building inspections, combined with a dashboard for managers to consume reports to determine whether further assessments or retrofits are required. The process of school building inspections was digitalized, with a chatbot implemented that features notifications either according to a routine inspection schedule or postseismic events, a conversation-based interface for guiding nonprofessional assessors, the integration of intuitive activation of inspections after receiving notifications, the use of multimedia to show damage directly without the possibility of mistakes, and data visualization for supporting managerial decision-making to enhance the quality and accuracy of budget allocation.
This research aims to build a Mandarin named entity recognition (NER) module using transfer learning to facilitate damage information gathering and analysis in disaster management. The hybrid NER approach proposed in this research includes three modules: (1) data augmentation, which constructs a concise data set for disaster management; (2) reference model, which utilizes the bidirectional long short-term memory–conditional random field framework to implement NER; and (3) the augmented model built by integrating the first two modules via cross-domain transfer with disparate label sets. Through the combination of established rules and learned sentence patterns, the hybrid approach performs well in NER tasks for disaster management and recognizes unfamiliar words successfully. This research applied the proposed NER module to disaster management. In the application, we favorably handled the NER tasks of our related work and achieved our desired outcomes. Through proper transfer, the results of this work can be extended to other fields and consequently bring valuable advantages in diverse applications.
This research aims to develop a knowledge base for a disaster management question-answering dialogue system. The rapid growth of the amount of data has led to the variance of data in terms of their formats, sources, and attributes. Hence, the difficulties of decision makers to accomplish their missions accurately and efficiently have increased. To solve this problem, we developed a questionanswering dialogue system for disaster management. In our previous research, we found that the information most likely retrieved in response to a user's request can be determined by calculating the similarity between the keywords and the user's input in a handcrafted keyword-information mapping table. However, we also noticed that managing the mapping table was a tedious task. Moreover, for the inputs that had more than one keyword, the system was unable to provide integrated information. Therefore, we constructed a knowledge base to optimize the performance and maintainability of the system. To build the knowledge base for disaster management, we designed the domain model by performing an abstraction on the knowledge of professional information providers and the required data on disaster management, while considering their source, certainty, and spatiotemporal features. The query of requested information from the knowledge base is composed of mentioned entities in the user's input. For the dialogue system to recognize the entities, we applied entity recognition. The subtasks include segmentation, tagging, similarity calculation with the names of the entities in the knowledge base, and intent detection to determine the desired knowledge of the user.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.