., Helmi (2025) Focused Web Crawler Untuk Jurnal Informatika Menggunakan Algoritma Shark Search Dengan IndoBERT dan KNN. Bachelor thesis, Institut Teknologi Kalimantan.
![[img]](http://repository.itk.ac.id/style/images/fileicons/text.png) |
Text
11211043_cover.pdf Restricted to Registered users only until 4 October 2027. Download (305kB) | Request a copy |
|
Text
11211043_statement_of_aunthenticity.pdf Restricted to Registered users only until 4 October 2027. Download (393kB) | Request a copy |
|
Text
11211043_publishing_agreement.pdf Restricted to Registered users only until 4 October 2027. Download (440kB) | Request a copy |
|
Text
11211043_approval_sheet.pdf Restricted to Registered users only until 4 October 2027. Download (399kB) | Request a copy |
|
Text
11211043_preface.pdf Restricted to Registered users only until 4 October 2027. Download (627kB) | Request a copy |
|
Text
11211043_abstract_id.pdf Restricted to Registered users only until 4 October 2027. Download (278kB) | Request a copy |
|
Text
11211043_abstract_en.pdf Restricted to Repository staff only until 4 October 2027. Download (276kB) | Request a copy |
|
Text
11211043_table_of_content.pdf Restricted to Repository staff only until 4 October 2027. Download (297kB) | Request a copy |
|
Text
11211043_ilustrations.pdf Restricted to Repository staff only until 4 October 2027. Download (251kB) | Request a copy |
|
Text
11211043_tables.pdf Restricted to Repository staff only until 4 October 2027. Download (227kB) | Request a copy |
|
Text
11211043_chapter_1.pdf Restricted to Repository staff only until 4 October 2027. Download (383kB) | Request a copy |
|
Text
11211043_chapter_2.pdf Restricted to Repository staff only until 4 October 2027. Download (981kB) | Request a copy |
|
Text
11211043_chapter_3.pdf Restricted to Repository staff only until 4 October 2027. Download (306kB) | Request a copy |
|
Text
11211043_chapter_4.pdf Restricted to Repository staff only until 4 October 2027. Download (1MB) | Request a copy |
|
Text
11211043_conclusions.pdf Restricted to Repository staff only until 4 October 2027. Download (256kB) | Request a copy |
|
Text
11211043_bibliography.pdf Restricted to Registered users only until 4 October 2027. Download (229kB) | Request a copy |
|
Text
11211043_enclosure.pdf Restricted to Repository staff only until 4 October 2027. Download (221kB) | Request a copy |
|
Text
11211043_paper.pdf Restricted to Repository staff only until 4 October 2027. Download (544kB) | Request a copy |
|
Text
11211043_paper.pdf Restricted to Repository staff only until 4 October 2027. Download (544kB) | Request a copy |
|
Text
11211043_Form. TA-020.pdf Restricted to Repository staff only until 4 October 2027. Download (740kB) | Request a copy |
|
Text
11211043_presentation.pdf Restricted to Repository staff only until 4 October 2027. Download (1MB) | Request a copy |
Abstract
Penelitian ini bertujuan untuk mengembangkan sistem focused crawler untuk pencarian jurnal informatika, memanfaatkan algoritma Shark Search yang dioptimalkan untuk melakukan pencarian sekaligus menelusuri halaman yang akan dikunjungi atau biasa disebut dengan crawling, serta memanfaatkan model bahasa IndoBERT dan algoritma K-Nearest Neighbors (KNN). Fokus utama penelitian ini adalah untuk meningkatkan harvest rate dan akurasi pengenalan dokumen yang relevan dengan topik informatika. Sistem dirancang untuk mengumpulkan dokumen secara efisien dari internet dengan memanfaatkan seed URLs dari Google Scholar, serta mengevaluasi kinerja model dengan menggunakan metrik evaluasi, seperti confusion matrix, akurasi, dan fi-score. Hasil penelitian menunjukkan bahwa integrasi IndoBERT dan KNN mampu menghasilkan akurasi hingga 94,81% dan f1-score sebesar 94,62%, dengan performa terbaik pada parameter K = 4. Selain itu, pengujian parameter Shark Search menghasilkan konfigurasi optimal pada δ = 0.4, β = 0.5, dan γ = 0.5 yang memberikan harvest rate sebesar 92,89%. Dibandingkan dengan algoritma BFS, Shark Search terbukti lebih unggul dan stabil dalam menjaga harvest rate di atas 90%, bahkan saat jumlah halaman yang di-crawl meningkat. Temuan ini menegaskan bahwa pendekatan selektif berbasis heuristik pada Shark Search lebih efektif dalam mengidentifikasi dokumen jurnal informatika yang relevan dibandingkan metode crawling konvensional.
| Item Type: | Thesis (Bachelor) |
|---|---|
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
| Divisions: | Jurusan Matematika dan Teknologi Informasi > Informatika |
| Depositing User: | Helmi . |
| Date Deposited: | 11 Jul 2025 08:09 |
| Last Modified: | 11 Jul 2025 08:09 |
| URI: | http://repository.itk.ac.id/id/eprint/23983 |
Actions (login required)
 |
View Item |