The purpose of this study is to produce a virtual teacher application using combination of natural language processing (NLP) and pattern matching methods to assist students when learning by themselves without being accompanied by a teacher. The author uses the research object in Domenico Savio Junior High School. Author’s research subjects are 7th, 8th and 9th grade students there. The author used natural language processing methods as a solution for application to conduct teaching and learning process with students using Indonesian language. Coupled with the pattern matching method, virtual teachers can classify the types of questions that students ask. This classification aims to solve problems that seem the same but have different solutions. The results of this study is virtual teacher application with facilities of creating material, making questions, and answering questions if that question material is contained in the application database. Domenico Savio junior high school teachers also can add or modify this knowledge