IMPLEMENTASI METODE TF – IDF (TEXT MINING) PADA APLIKASI CHATBOT UNTUK TANYA JAWAB TENTANG FAQ PENDAFTARAN MAHASISWA BARU STT-PLN
Total View This Week0
Total View This Week0
Institusion
Institut Teknologi Perusahaan Listrik Negara
Author
WIDIA, NURUL
Abdurrasyid, Abdurrasyid
Widiyanto, Max Teja Ajie Cipta
Abdurrasyid, Abdurrasyid
Widiyanto, Max Teja Ajie Cipta
Subject
Teknik Informatika
Datestamp
2022-09-23 02:19:53
Abstract :
Chatbot is an application designed to communicate with the machine to acquire informations. Informations obtained are able to hopefully help answering the questions regarding the problems faced by students about The New Students Admission of STT-PLN without place and time limits and can also be used time and time again.The chatbot is built by applying TF - IDF (Text Mining) and Cosine Similarity method. TF - IDF method is a means of giving relation significance of a term towards the question. Cosine similarity method is a method to calculate the similarity between two objects expressed by two vectors using keywords from a question as a measurement. The system development method used in this thesis is waterfall model. Training data for this research are as many as 8 questions FAQ. The result of this research shows system classification accuracy level percentage amounting to 98%
Chatbot is an application designed to communicate with the machine to acquire informations. Informations obtained are able to hopefully help answering the questions regarding the problems faced by students about The New Students Admission of STT-PLN without place and time limits and can also be used time and time again.The chatbot is built by applying TF - IDF (Text Mining) and Cosine Similarity method. TF - IDF method is a means of giving relation significance of a term towards the question. Cosine similarity method is a method to calculate the similarity between two objects expressed by two vectors using keywords from a question as a measurement. The system development method used in this thesis is waterfall model. Training data for this research are as many as 8 questions FAQ. The result of this research shows system classification accuracy level percentage amounting to 98%
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
