SISTEM TANYA JAWAB UNTUK PERTANYAAN APA MENGGUNAKAN METODE PENDEKATAN JARAK

GALEA, ZAFIRA and Yusliani, Novi (2023) SISTEM TANYA JAWAB UNTUK PERTANYAAN APA MENGGUNAKAN METODE PENDEKATAN JARAK. Undergraduate thesis, Sriwijaya University.

	Text RAMA_55201_09021281924049.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (23MB)
	Text RAMA_55201_09021281924049_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (5MB) | Request a copy
	Text RAMA_55201_09021281924049_0008118205_01_front_ref.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (4MB) | Request a copy
	Text RAMA_55201_09021281924049_0008118205_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (4MB) | Request a copy
	Text RAMA_55201_09021281924049_0008118205_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (4MB) | Request a copy
	Text RAMA_55201_09021281924049_0008118205_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (2MB) | Request a copy
	Text RAMA_55201_09021281924049_0008118205_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (3MB) | Request a copy
	Text RAMA_55201_09021281924049_0008118205_06.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (412kB) | Request a copy
	Text RAMA_55201_09021281924049_0008118205_07_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_55201_09021281924049_0008118205_08_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (2MB) | Request a copy

Abstract

Question Answering System (QAS) is a system designed to generate information quickly in the form of natural language. The information produced by this system consists of short text snippets as answers to user questions. In this research on Question Answering System (QAS), users can obtain answers related to non-factoid information questions. The generated information is close domain, sourced from a segment of biology learning information obtained from the best documents. The information retrieval process involves preprocessing to obtain keywords and important words, calculating TF-IDF, and using three distance-approach algorithms: Manhattan Distance, Euclidean Distance, and Cosine Similarity to retrieve the best documents and answers. In the testing phase, 110 question-answer pairs and 61 documents containing answers were used. The best accuracy was achieved by the Manhattan Distance algorithm, with 0.9% in test 1, where only the definition answers were considered. The best accuracy for Euclidean Distance was 36.36%, and for Cosine Similarity, it was 37.27% in test 4, where both the definition answers and keywordswith a 40% threshold were considered, obtaining a reasonably good result. This research demonstrates the capability to perform question-answering for non-factoid questions, and it highlights how the preprocessing process, the type of distance-approach algorithms used, and the testing methods can affect the accuracy of the system. Keywords: QAS, Manhattan Distance, Euclidean Distance, Cosine Similarity.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	QAS, Manhattan Distance, Euclidean Distance, Cosine Similarity.
Subjects:	T Technology > T Technology (General) > T10.5-11.9 Communication of technical information
Divisions:	09-Faculty of Computer Science > 55201-Informatics (S1)
Depositing User:	Zafira Galea
Date Deposited:	03 Aug 2023 08:14
Last Modified:	03 Aug 2023 08:14
URI:	http://repository.unsri.ac.id/id/eprint/125407

Actions (login required)

View Item