A User-Centered Design of Natural Language Processing for Maternal Monitoring Chatbot System
A self-monitoring device and remote counseling system for pregnant mothers using interactive chat is an effective method due to the reduction of maternal care visits caused during the pandemic. The employment of an artificial intelligence (AI) based system using natural language processing (NLP)...
| Main Authors: | , , , , , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English |
| Published: |
2022
|
| Subjects: | |
| Online Access: | https://repository.ugm.ac.id/278963/1/Afrizal_KKMK.pdf |
| Summary: | A self-monitoring device and remote counseling system for pregnant mothers using interactive chat is an effective method due to the reduction of maternal care visits
caused during the pandemic. The employment of an artificial
intelligence (AI) based system using natural language
processing (NLP) for decision support has prospectively
enhanced the conversational access of patient to improve health awareness and knowledge. This research was conducted to develop an AI-based system which focuses on education and
monitoring with regard to danger signs during pregnancy using NLP. The Telegram chatbot was used to develop the system after investigating user needs based on the danger sign monitoring guideline from WHO and the Ministry of Health of the Republic of Indonesia. The inputs from users were recognized by NLP and forwarded to the testing data for
decision system. Furthermore, the analysis result was sent to the user which provides educational information and a personalized monitoring result. System Usability Scale (SUS) was undertaken to assess the user ability to use the application. The SUS score average for the chatbot system was 62.3 which was classified as “OK” for the adjective ratings. The implication of the maternal monitoring using a chatbot system is the improvement of maternal care with regard to early detection of danger signs and relevant suggestions using an effective and interactive system
which could be very promising especially in a limited healthcare resource environment. |
|---|