Please use this identifier to cite or link to this item:
https://repositori.uma.ac.id/handle/123456789/24214
Title: | Analisis Metode Clustering dengan Algoritma Spectral Clustering dalam Pengelompokan Tingkat Tindak Kriminalitas |
Other Titles: | Analysis of the Clustering Method with the Spectral Clustering Algorithm in Grouping Crime Levels |
Authors: | Pratama, Egi |
metadata.dc.contributor.advisor: | Rahmadsyah |
Keywords: | Crime;Data Mining;Clustering;Spectral Clustering;Kriminalitas |
Issue Date: | 21-Dec-2023 |
Publisher: | Universitas Medan Area |
Series/Report no.: | NPM;198160027 |
Abstract: | Kriminalitas merupakan tindakan yang menggangu ketertiban dan kenyamanan umum, kriminalitas dianggap sebagai tindakan yang negatif karena merugikan korban baik secara fisik ataupun secara mental. Di Indonesia sendiri terdapat berbagai macam tindakan kriminalitas, mulai dari tindakan yang sangat berbahaya seperti pembunuhan, pencabulan. Kemudian tindakan yang dianggap berbahaya seperti pencurian, serta tindakan yang di anggap cukup berbahaya yaitu tindakan kejahatan seperti kerusuhan umum, demonstrasi. Tindakan kriminalitas yang meradang perlu untuk dianalisis untuk kemudian dapat dicegah sesuai dengan kapasitas hukum dan tindakan pencegahan. Penelitian ini menggunakan data dari masyarakat indonesia yang mengeluh di tweeter tentang tindakan kriminalitas yang marak terjadi, menggunakan 10.000 sempel data yang dikumpulkan pada tanggal 25 Juni 2023 menggunakan teknik data mining sebagai bahan analisis dan menggunakan metode spectral clustering sebagai algoritma analisis untuk mendapatkan hasil klasterisasi tindakan kejahatan. Data dianalisis menjadi tiga kluster yaitu kluster sangat berbahaya dengan presentasi 90.3% atau 9026 data, kemudian kluster berbahaya dengan presentasi 7% atau 697 data, dan kluster cukup berbahaya dengan presentasi 2.8% atau 277 data. Crime is an act that disturbs public order and comfort, crime is considered a negative action because it harms the victim both physically and mentally. In Indonesia itself there are various kinds of criminal acts, ranging from very dangerous acts such as murder, obscenity. Then actions that are considered dangerous, such as theft. As well as actions that are considered quite dangerous, namely acts of crime such as public riots, demonstrations. Inflammatory criminal acts need to be analyzed so that they can be prevented in accordance with legal capacity and preventive measures. This study uses data from Indonesian people who complain on tweeters about criminal acts that are rife, uses 10,000 samples of data collected on June 25, 2023 using data mining techniques as material for analysis and uses the spectral clustering method as an analytical algorithm to get the results of clustering crimes. The data were analyzed into three clusters, namely the very dangerous cluster with a presentation of 90.3 % or 9026 data, then the dangerous cluster with a presentation of 7% or 697 data, and the moderately dangerous cluster with a presentation of 2.8% or 277 data. |
Description: | 88 Halaman |
URI: | https://repositori.uma.ac.id/handle/123456789/24214 |
Appears in Collections: | SP - Informatic Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
198160027 - Egi Pratama - Chapter IV.pdf Restricted Access | Chapter IV | 567.92 kB | Adobe PDF | View/Open Request a copy |
198160027 - Egi Pratama - Fulltext.pdf | Cover, Abstract, Chapter I, II, III, V, Bibliography | 1.4 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.