Intelligent System for Agent in Educational Game Using Dynamic Gram Similarity

• Published in: 2018 International Seminar on Application for Technology of Information and Communication pp. 373 - 376
• Main Authors: Ampuh Yunanto, Andhik, Herumurti, Darlis, Kuswadayan, Imam, Rochimah, Siti
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.09.2018
• Subjects: Artificial Intelligent; Computational modeling; Educational Game; Games; Grammar; Grammar English; Information retrieval; Intelligent systems; Natural Language Processing; Seminars; Text Similarity
• DOI: 10.1109/ISEMANTIC.2018.8549718

Today, almost people like a game. A game can make people have fun when playing it. An exciting game must have many feature and menus especially artificial intelligent (AI) called enemy or agent. Agent in game can be build using any method to make smart and intellect. In educational game, smart agent must can choose and decide a right answer of question. In this research, we make an agent to answer English question using dynamic gram similarity. This method based on n-gram similarity method but it can check neighbor word on sentence. Result of this research indicates dynamic gram similarity has 66% which is more accurate answer than unigram, big ram and trigram similarity. The conclusion of this research is dynamic gram similarity can implement to agent in educational game and can answer grammar English question automatically.