Hasil Pencarian  ::  Simpan CSV :: Kembali

"ABSTRACTCriminality is a social problem causing negative impacts on society welfare. Police as law enforcement officer was required to take actions to prevent criminality which was increasingly widespread. Such efforts could be realized by analizing criminal data to obtain useful information for the preparation of criminal prevention strategies. However, extracting knowledge from criminal data effectively was a problematique for them. In this study, data mining was used to solve knowledge extraction problem from the dataset. The technique was aimed to get information about crime patternsby analyzing criminal activity habits. Association rule mining and apriori algorithm were used to find crime patterns. Generating crime patterns in data mining was difficult to understand when there were too many rules. Graph based visualization of association rules designed to solve that problem. Generated visualization showed relationship between crimes. That visualization was expected to help the police to understand the crime pattern so they could do prevention efforts more effectively. The results showed that the visualization of association rules could present association rules in more interesting way and described the crime pattern."

"Human Immunodeficiency Virus (HIV) merupakan virus yang menyerang sistem kekebalan tubuh manusia. Virus ini terdiri dari 23 protein dalam RNA untai tunggal. Interaksi protein HIV dan protein manusia dapat mengakibatkan penyakit AIDS. Dengan mempelajari interaksi protein dapat digunakan untuk mengembangkan obat antiviral. Untuk menganalisis interaksi protein dilakukan dengan proses biclustering. Algoritma LCM-MBC merupakan suatu algoritma biclustering yang digunakan untuk menganalisis interaksi protein. Hasil dari biclustering digunakan untuk memprediksi dengan association rule mining. Untuk mengetahui fungsi-fungsi biologis dari protein yang terdapat pada satu bicluster digunakan DAVID Gene Ontology. Terdapat 45 bicluster yang memiliki protein HIV dalam satu bicluster sebanyak lima. Dari bicluster yang diperoleh ini, Terdapat 11 protein HIV-1 yang diprediksi akan berinteraksi dengan 36 protein manusia. Jika protein manusia terhubung dengan protein HIV sesuai dengan tipe jenis interaksinya, artinya protein manusia tersebut berinteraksi dengan proten HIV-1.

---

Human Immunodeficiency Virus (HIV) is a virus w attacks the human immune system. This virus consists of 23 proteins in a single-stranded RNA. The protein interaction between HIV proteins and human proteins can impact to AIDS The research about HIV-1 proteins and human proteins interactions leads to the insight of drug target prediction. To analyze protein interactions carried out by biclustering process. The LCM-MBC algorithm is a biclustering algorithm that is used to analyze protein interactions.

The results of biclustering are used to predict with association rule mining. To find out the biological functions of proteins found in one cluster used DAVID Gene Ontology. There are 45 bicluster that have five HIV proteins in one bicluster. From the bicluster obtained, there are 11 HIV-1 proteins that are predicted to interact with 36 human proteins. If human protein interacts with HIV-1 protens, it means that human proteins will relate according to the interaction type by HIV proteins."

"Industri Pakaian di Indonesia telah berkembang dengan sangat cepat. Peningkatan penggunaan ecommerce di bidang fashion telah menghasilkan kompetisi yang tinggi antar brand global dan brand local. Oleh karena itu diperlukanlah strategi pemasaran yang penting dan baik untuk menjaga pertumbuhan industri lokal. Penelitian ini bertujuan untuk menggunakan RFM model dan Association Rule Mining (ARM) untuk membantu mengetahui segmentasi konsumen. ARM adalah salah satu Teknik paling popular untuk mengetahui pola dari atribut – atribut yang ada pada database, dan RFM model digunakan untuk mengetahui perilaku konsumen. Setelah data dikumpulkan, dilakukan preprocessing, dan dilakukan analisis RFMnya, kemudian dilakukan k-means clustering. Setelah ditemukan cluster dari konsumen, dilakukan ARM untuk mencari pola dari tipe konsumen, promosi diskon dan promosi ongkos kirim yang mereka pakai. Kemudian disusun profil konsumen berdasarkan pola dan nilai RFM konsumen yang didapatkan.

---

The apparel and fashion industry of local brands in Indonesia has been growing rapidly. A good and strategic marketing strategy is needed to maintain the industry growth and sustain the local industry. This research aims to build marketing strategy with segmentation. The study utilized the Machine Learning using Association Rules Mining (ARM) and Consumer segmentation of RFM model. The ARM is one of the most popular techniques to learn a pattern or associations of attributes of customers. Consumer segmentation of RFM models was used to understand about consumer’s behaviors. The data was collected from a local fashion brands in e- commerce platforms. After data was collected, the data was preprocessed, and then analyzed using RFM model. After the RFM model concluded, the model was used to associate consumers type, discount and delivery promotion by using ARM to understand about the relations about the promotions and the consumers type. The segmentation is done by clustering with k-means algorithm. After customer segmentation concluded, the marketing strategy is then built with a marketing mix approach"

"Peningkatan persaingan dan kunjungan di situs web e-commerce shopping mall di Indonesia perlu disertai dengan meningkatkan strategi Customer Relationship Management CRM . Strategi yang bisa digunakan adalah peningkatan kualitas pelayanan, hal ini bisa di implementasikan melalui penyusunan sistem rekomendasi produk di situs web e-commerce tersebut. Untuk menyusun sistem tersebut, penggalian pola asosiasi produk dilakukan dengan memanfaatkan data web log yang berisi data navigasi dan pola kebiasaan pelanggan. Hal tersebut diakomodasi oleh metode web usage mining yaitu association rules. Algoritma yang digunakan adalah algoritma yang memberikan input asosiasi berdasarkan frekuensi item, yakni algoritma Apriori. Untuk menguji dan menyeleksi pola yang dihasilkan, objective interestingness measure dilakukan dan menghasilkan 25 luaran pola asosiasi.

---

An increasing of competition and visitors on e commerce shopping mall websites in Indonesia, need to be accompanied by improving Customer Relationship Management strategy. A strategy that can be used is improving the quality of services, it can be implemented through the preparation of product recommendation system on the e commerce website. To compile the system, pattern recognition of product association is conducted by utilizing weblog data which contains navigation data and customer behavior pattern. It is accommodated by web usage mining method that is association rules. The algorithm applied is an algorithm that provides input association based on item frequency, i.e Apriori algorithm. To test and select the resulting pattern, objective interestingness measure was performed and yields 25 outcomes of the association pattern."

"Online customers segmentation could be a valuable research topic of marketing strategy. Previous literature mainly studied the differences between non-purchasers and purchasers, lacking further segmentation of online customers themselves. This thesis focuses on online customer segmentation based on a large volume of real transaction data in one of Indonesias e-commerce website. This research proposes a customer clustering technique using the K-Means algorithm and RFM Patterns as an analysis of the customers profile. Then, the market basket analysis is conducted using the Apriori algorithm for every customer profile and cluster to obtain the association rule as well as product relationships purchased by customers. Later on, the result of market basket analysis is utilized as an input for e-commerce companies in designing promotions such as bundling or product recommendation system for segmented customers.

---

Segmentasi pelanggan daring bisa menjadi topik penelitian yang berharga dalam strategi pemasaran. Literatur yang sudah ada cenderung mempelajari perbedaan antara pembeli dan non-pembeli, tanpa menggali lebih lanjut mengenai segmentasi pelanggan daring itu sendiri. Tesis ini berfokus pada segmentasi pelanggan daring berdasarkan data transaksi di salah satu situs penjualan daring di Indonesia. Penelitian ini mengusulkan teknik pengelompokan pelanggan menggunakan algoritma K-Means dan pola RFM sebagai analisis profil pelanggan. Kemudian, analisis keranjang belanja dilakukan dengan menggunakan algoritma Apriori untuk setiap profil pelanggan dan kluster untuk mendapatkan aturan asosiasi serta hubungan produk yang dibeli oleh pelanggan. Kemudian, hasil analisis keranjang belanja tersebut digunakan sebagai masukan untuk perusahaan penjualan daring dalam merancang promosi seperti bundling atau sistem rekomendasi produk untuk pelanggan yang berada dalam profil yang sama."

"ABSTRAKLibrary X is an academic library in Jakarta, Indonesia. Library X has provided Online Public Access Catalog (OPAC) as a tool to provide information related to the collection. However, sometimes the information appears does not show high relevancy. One way to solve this problem is to develop user need based-book recommendation system. The purpose of this study is to create personalization model of book recommendations in Library X.Data Collection Method. The method used in this study was association rule mining using Apriori algorithm. Results and Discussions. The results showed that the book relationships for the minimum support was 0.1% and the minimum confidence was 10% and generated 42 association rules. It is noted that 657 (Accounting) and 658 (Management) are found to support for 2.6% with the confidence level for 14%.Conclusions. Book recommendation is formulated by selecting the rule with maximum support and confidence. The recommendation system is designed to be integrated to web application and users e-mail."

"Association rules are used to predict frequent web user behaviors from web usage data. These rules are formed using frequent items. The number of association rules increases as the number of frequent items increases and produces several redundant and non-informative rules. In this paper, five interestingness measures, including cosine, lift, leverage, confidence, and conviction with a constant value of support are compared based on the number of redundant and non-informative rules that they produce. Redundant and non-informative rules are a subset of rules present in the top generated rules. The experimental results suggested that leverage produced the least number of redundant rules in the top rules but also produced the least informative rules among all measures. Lift showed the highest number of redundant rules but the most informative rules among all the measures."

"Industri konstruksi merupakan industri dengan tingkat risiko tinggi dan menjadi industri paling berbahaya di seluruh dunia. Hal ini mendorong kebutuhan adanya sistem kontrol dan upaya pencegahan keselamatan yang efektif, terutama dalam mengidentifikasi bahaya melalui proses Learning From Incidents. Penggunaan data mining dalam keselamatan konstruksi mulai banyak digunakan dalam penelitian. Namun, diperlukan model yang dapat membantu praktisi mengembangkan data mining untuk mengidentifikasi bahaya di proyek konstruksi. CRISP-DM sebagai standar de facto model data mining dapat diimplementasikan untuk menjadi standar dan pedoman bagi praktisi. Tujuan utama penelitian ini adalah mengembangkan model CRISP-DM untuk improvement proses dalam mengidentifikasi bahaya proyek konstruksi serta memperoleh Learning From Incidents Database yang terbentuk dengan studi kasus pekerjaan proyek konstruksi struktur atas jalan layang beton. Association Rule Mining menjadi metode data mining yang digunakan dalam penelitian ini untuk mendapatkan aturan asosiasi antara aktivitas pekerjaan dan bahaya yang terjadi. Hasil penelitian menunjukkan terdapat langkah-langkah praktis yang dapat dilakukan untuk mengembangkan model CRISP-DM dalam identifikasi bahaya konstruksi. Percobaan implementasi CRISP-DM tersebut menghasilkan database yang menunjukan 5 aturan asosiasi dengan rerata akurasi 51,2% dari 112 kejadian kecelakaan konstruksi di Indonesia. Pakar keselamatan konstruksi juga menilai database dari aturan asosiasi yang terbentuk telah sesuai dengan kondisi aktual secara umum dan model CRISP-DM yang diajukan dapat meningkatkan Learning From Incidents pada industri konstruksi. Namun, peningkatan sistem pelaporan, investigasi, serta kesadaran pentingnya keselamatan masih perlu ditingkatkan sebelum model CRISP-DM dapat diterapkan di industri konstruksi Indonesia.

---

The construction industry is a high-risk industry and the most dangerous industry in the world. This drives the need for hazard identification through the Learning from Incidents process. The use of data mining in construction safety is starting to be widely used in research. However, a model is needed that can help practitioners develop data mining to identify hazards in construction projects. The main objective of this research is to develop a CRISP-DM model for process improvement in identifying construction project hazards and obtain a Learning from Incidents Database formed with a case study of concrete elevated road structure construction work. The data mining method used in this research is Association Rule Mining. The results showed that there are practical steps that can be taken to develop the CRISP-DM model. The implementation of the model produced 5 association rules with an average accuracy of 51.2% of 112 construction accidents. Experts assessed that the association rules formed are in accordance with the actual conditions and the CRISP-DM model can improve Learning from Incidents in the construction industry. However, improvements in reporting systems, investigations, and safety awareness still need to be improved before the model can be applied in Indonesian construction industry."

"Industri otomotif memberikan konstribusi terbesar dalam produk domestik bruto Indonesia. Seiring dengan semakin pesatnya perkembangan industri otomotif, terutama industri mobil, persaingan antar produsen mobil semakin tinggi. Hal ini mengakibatkan hasil keluaran produk dari berbagai produsen mobil akan mencapai tahap dimana hampir semuanya memiliki standar kualitas yang sama. Keadaan ini mendorong konsumen untuk memanfaatkan faktor lain di luar spesifikasi fungsional dan kualitas untuk membuat keputusan membeli mobil, yaitu persepsi afektif. Penelitian ini membahas bagaimana konsumen mobil tipe city car di Indonesia menilai produk dari sisi bentuk eksterior dengan mengeluarkan sisi afektif. Metode yang digunakan dalam penelitian ini adalah Kansei Engineering, lebih tepatnya menggunakan Kansei Words. Pengolahan data dari konsumen dilakukan dengan metode association rule mining dan conjoint analysis. Dari hasil peneltiian, diketahui bahwa terdapat lima kelompok Kansei Words yang mewakili persepsi afektif konsumen, yaitu klasik (classic) dan ramping (sleek), mantap/kuat (robust) dan bertenaga/tangguh (powerful), mencolok (sporty) dan formal (formal), lucu/imut (cute), dan modern (modern). Keluaran akhir dari penelitian ini adalah terbentuknya 5 usulan desain baru untuk bentuk eksterior city car yang memenuhi masing-masing kelompok Kansei Words di atas.

---

Automotive industry delivering a great contribution to Indonesia by accounting high percentage in gross domestic product. As automotive industry is developing, especially for car industry, the competition between car companies is highly increasing. This condition resulted in a situation where products from different car companies having the same standard for quality. Therefore, customers are triggered to consider another factor beside functional specification and quality, which is affective perception.

This research focused on how customers of city car in Indonesia evaluate the product from its exterior shape by considering their affective side. Method of this research is Kansei Engineering, specifically its Kansei Words. Data from customers are processed with the method of association rule mining and conjoint analysis.

From the output of this research, there are 5 groups of Kansei Words that represent customers affective perception, which are classic and sleek, robust and powerful, sporty and formal, cute, and modern. The final output from this research are 5 recommended designs for city car exterior shape that describe all the Kansei Words above."