Chit Chat Cinema: Question Answering System dengan Domain Film Bioskop

Setiawan, Ryan Agus (2014) Chit Chat Cinema: Question Answering System dengan Domain Film Bioskop. Other thesis, Sekolah Tinggi Teknik Musi.

	Text IF-2014-310003-cover.pdf Download (380kB)
	Text IF-2014-310003-abstract.pdf Download (202kB)
	Text IF-2014-310003-tableofcontent.pdf Download (441kB)
	Text IF-2014-310003-chapter1.pdf Download (442kB)
	Text IF-2014-310003-chapter2.pdf Restricted to Repository staff only Download (1MB)
	Text IF-2014-310003-chapter3.pdf Restricted to Repository staff only Download (4MB)
	Text IF-2014-310003-chapter4.pdf Restricted to Repository staff only Download (4MB)
	Text IF-2014-310003-conclusion.pdf Download (88kB)
	Text IF-2014-310003-reference.pdf Download (508kB)
	Text IF-2014-310003-attachment.pdf Restricted to Repository staff only Download (3MB)
	Text IF-2014-310003-complete.pdf Restricted to Repository staff only Download (4MB)
	Text IF-2014-310003-summary_id.pdf Restricted to Repository staff only Download (458kB)

Abstract

Along with the development of times, informations stored on Internet will spread more widely. However, it is not matched by the ease in finding desired information. Users need to browse web pages for further information desired. Several sites also displays a lot of images that resulted in long waiting times, big internet traffic consumption, heavy and impressed when accessed via mobile phone. Based on these problems, then application named Chit Chat Cinema was built. This application is an implementation of the Question Answering System where users simply enter question by using Indonesian language and it will give an appropriate answer. Questions that can be recognized is the title of movie based on a particular actor, now playing movie, coming soon movie, duration, actor's name whose played by particular movie title, genre, and synopsis. Web service that located on server created using NodeJS and MySQL database. Question processing methods are tokenization, part-of-speech tagging, and rule-based named entity recognition. Process model is sequential linear. From 260 questions were tested, there are 208 questions can answered correctly. The system gives wrong answer if question use unknown pattern, contain words that aren’t recognized, and the name of entity is an Indonesian loan word.

Actions (login required)

View Item