Handling Unanswerable Question in Machine Reading Comprehension by using Sentence Transformers - Dalam bentuk buku karya ilmiah

HADDAD ALWI YAFIE

Informasi Dasar

72 kali
23.05.347
006.35
Karya Ilmiah - Thesis (S2) - Reference

Machine Reading Comprehension (MRC) is a fundamental task within question-answering systems, aimed at generating context-aligned answers to given questions. Traditionally, MRC systems focus on providing answers to posed questions. However, contemporary research explores the realm of unanswerable questions, where responses genuinely acknowledge the absence of an answer. This research its efforts on addressing unanswerable questions using a retriever-reader approach, leveraging a sentence transformer for semantic-based document retrieval. Evaluation is performed through metrics like Exact Match and F1-score, showcasing significant performance improvements compared to previous retriever-reader methods. Additionally, the sentence transformer proves superior in document and context retrieval compared to the traditional Tf-Idf approach. The proposed method successfully answers questions that have no answers or unanswerable questions with an exact match rate of 89% and an F1-Score of 89% for the unanswerable dataset.

Subjek

NATURAL LANGUAGE PROCESSING
 

Katalog

Handling Unanswerable Question in Machine Reading Comprehension by using Sentence Transformers - Dalam bentuk buku karya ilmiah
 
xiv, 36p.: il ,;pdf file
English

Sirkulasi

Rp. 0
Rp. 0
Tidak

Pengarang

HADDAD ALWI YAFIE
Perorangan
Arie Ardiyanti Suryani
 

Penerbit

Universitas Telkom, S2 Informatika
Bandung
2023

Koleksi

Kompetensi

 

Download / Flippingbook

 

Ulasan

Belum ada ulasan yang diberikan
anda harus sign-in untuk memberikan ulasan ke katalog ini