<![CDATA[The growth in the number of movies continues to be experienced every year, making it difficult for users to choose the right movie. The recommender system is an alternative to being able to solve the problem. In many studies, recommender systems have been developed, but in their use, they do not apply intensive interaction between users and the system created. In this study, we developed a chatbot to help implement a movie recommender system that ensures users can interact intensively with the system with natural language. The chatbot was created using Dialog low to enable the system to recognize the natural language. One way to understand a text concept is to find the relationship between the text concepts. Latent Semantic Analysis (LSA) can implement this, where LSA has the advantage of extracting a text and making a statistical representation in the form of a term-document matrix (TF-IDF) using a lower dimension (low-rank approximation). Singular Value Decomposition (SVD) can help decompose a large matrix into a matrix with small dimensions to determine a text's overall meaning. The relationship between text concepts with the highest or almost the same probability value can be used as an output to respond to the user. From the test results, the chatbot shows that the match rate between system and user responses is 86%. Thus, the developed chatbot can be used well in providing interactive movie recommendations.]]>
Copyrights © 2022
