SISTEM CHATBOT SEBAGAI BANTUAN LAYANAN INFORMASI AKADEMIK MENGGUNAKAN FRAMEWORK RASA OPEN SOURCE DI UPN "VETERAN" JAKARTA

• Main Author: Bagas Prabaswara Wira Notokoesoemo, (Author)
• Format: Book
• Published: 2023-07-03.
• Subjects: Thesis; NonPeerReviewed
• Online Access: Link Metadata
• Notes: http://repository.upnvj.ac.id/25035/27/ABSTRAK.pdf; http://repository.upnvj.ac.id/25035/16/AWAL.pdf; http://repository.upnvj.ac.id/25035/17/BAB%201.pdf; http://repository.upnvj.ac.id/25035/18/BAB%202.pdf; http://repository.upnvj.ac.id/25035/19/BAB%203.pdf; http://repository.upnvj.ac.id/25035/20/BAB%204.pdf; http://repository.upnvj.ac.id/25035/21/BAB%205.pdf; http://repository.upnvj.ac.id/25035/22/DAFTAR%20PUSTAKA.pdf; http://repository.upnvj.ac.id/25035/23/RIWAYAT%20HIDUP.pdf; http://repository.upnvj.ac.id/25035/24/LAMPIRAN.pdf; http://repository.upnvj.ac.id/25035/11/HASIL%20PLAGIARISME.pdf; http://repository.upnvj.ac.id/25035/25/ARTIKEL%20KI.pdf
• Keywords: QA75 Electronic computers. Computer science; QA76 Computer software

The inefficiency of finding academic information at UPN "Veteran" Jakarta made all campus members such as students, lecturers, education staff and the general public choose to ask service officers directly or via chat. However, limited human resources as service officers make academic information services via chat not optimal. Therefore, this research developed a chatbot that can serve requests for information from students, lecturers, education staff and the general public autonomously through a messenger application and also developed a website intended for academic service officers so that they can interact with the chatbot's knowledge base if there are additions, changes or data deletion. This study uses data derived from the UPN Veteran Jakarta 2020/2021 academic handbook, data from interviews with the Integrated Services Unit and questionnaires to students, lecturers, UPN "Veteran" Jakarta education staff and the general public as training data for chatbots. Chatbot was developed using a machine learning framework, namely Open Source Sense. The results of the evaluation of data modeling and testing using Open Source Sense for nlu and dialogue management get an accuracy of 100% and 98%. While testing the functionality of the chatbot system function is functioning as expected. It can be said that chatbots built using Open Source Sense can increase the efficiency and productivity of academic information services for service officers and make it easier for students, lecturers, education staff and the general public to find information.