Claim Missing Document
Check
Articles

Found 7 Documents
Search
Journal : INFOKUM

IMPLEMENTATION OF TF-IDF AND COSINE SIMILARITY ALGORITHMS FOR CLASSIFICATION OF DOCUMENTS BASED ON ABSTRACT SCIENTIFIC JOURNALS Paska Marto Hasugian; Jonson Manurung; Logaraz Logaraz; Uzitha Ram
INFOKUM Vol. 9 No. 2, June (2021): Data Mining, Image Processing and artificial intelligence
Publisher : Sean Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (503.77 KB)

Abstract

Research on one of the higher education dharmas is carried out by each lecturer and is a challenge for lecturers who pay attention to produce new and useful findings. Research results will be published in journals both nationally and internationally and one of the websites published by Ristekbirn is Sinta which includes all research works in Indonesia. The problem in this research is the accumulation of data that is getting bigger and it needs to be analyzed by utilizing text mining by searching for the resources contained in the abstract document and presenting part of the information. The purpose of this study is to classify the suitability of another document so that knowledge is found. and placement in groups according to existing topics. The process of these problems is by classifying documents based on abstracts from the publication of scientific papers. Solving these problems involves two mutually supporting algorithms, namely TD-IDF with Cosine Similarity with different tasks. TF-IDF ensures the weight of each document that can be read and read with Cosine Similarity. This research uses text mining as part of the search for related patterns and documents that have been tested. For the process of calculating the test data, 1 document and 15 documents were used as training data. With the calculation of TD-IDF the weight of each document from Q, D2 to D15 is 10,946, 28,050,27,176, 39,043, 36,535, 30,696, 25,612, 12,581, 42,335, 29,661, 33,867, 31,706, 22,654, 15,450, 59,832, 42,127, The similarity of the data is tested by determining the value of k = 4 which results in similarity to the Expert System and Cryptography, while with the selection of K = 5 with the highest similarity to the expert system..
IMPLEMENTATION OF K-NEAREST NEIGHBOR ALGORITHM TO PERFORM CLASS PLACEMENT CLASSIFICATION AT GKPI PADANG BULAN JUNIOR HIGH SCHOOL Dewi Lasmiana Panjaitan; Paska Marto Hasugian
INFOKUM Vol. 10 No. 1 (2021): Desember, Data Mining, Image Processing, and artificial intelligence
Publisher : Sean Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (215.592 KB)

Abstract

Superior classes are a number of students who have outstanding abilities or achievements in these students, who are grouped in one particular class. One way that is done is the process of class placement. But at the time of class placement there are problems that arise, namely during the process of determining the class, whether students enter the superior class or ordinary classes. Students who have certain abilities will later occupy superior classes and students who do not have certain abilities do not enter the superior class. With this research will help the school in determining superior classes and ordinary classes, so that no one is harmed, which should be students who deserve to be superior classes. The purpose of this study is to implement the principle of data mining to class placement classification using the K-Nearest Neighbor Algorithm. Where the K-Nearest Neighbor Algorithm will classify objects based on learning data that is the closest to the object. Based on the results of the trial conducted by utilizing the K-NN algorithm with tested data as many as 64 data and training data as much as 82 data, it was obtained the results of class placement with students who occupy class A as many as 26 students, students who forged class B as many as 20 students and students who occupy class C as many as 18 students.
Determination Of The Best Private Universities Using The Analytical Hierarcy Process Method Pandi Barita Nauli Simangunsong; Paska Marto Hasugian
INFOKUM Vol. 9 No. 2, June (2021): Data Mining, Image Processing and artificial intelligence
Publisher : Sean Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (730.171 KB)

Abstract

Private Universities (PTS) are an option to continue post-graduate education, especially in the city of Medan. Quality private universities are very influential in carrying out education. To make it easier to compete in the world of work. Because a good PTS has a good accreditation, because PTS accreditation is very influential in the world of work. Especially in the career path, acceptance of new workers. Determining PTS in this Medan city is not easy, because there are many PTS in this Medan city. There are too many private universities in existence, making it difficult to choose PTS manually, and the results of the selection are sometimes inaccurate and become a problem in selecting the best PTS. Decision Support System (DSS) is a system that can assist a person in making accurate and targeted decisions. The method used in the Decision Support System is to use the Analytical Hierarchy Process (AHP). This method was chosen because it is able to find the best alternative from a number of alternatives, in this case the intended alternative is the one that has the right to become the best private university based on the specified criteria.
Decision Support System For Selection Of Electric Light Ball For Household With Technique Method For Order Preference By Simillarity To Ideal Solution (Topsis) Pandi Barita Nauli Simangunsong; Paska Marto Hasugian; Makmur Tarigan
INFOKUM Vol. 10 No. 1 (2021): Desember, Data Mining, Image Processing, and artificial intelligence
Publisher : Sean Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (435.898 KB)

Abstract

In the selection of these light bulbs, users are often confused with the choice of light bulbs that are so widely circulated in the market, ranging from brands, types, quality, prices that continue to compete and with other advantages on offer often make users confused to get Energy efficient light bulbs at affordable prices and with the best quality. The problem doesn't just stop there when placing light bulbs in each room of the house also often gets into trouble. The problem that is often caused is mismatch, this usually happens when the light bulb has been placed. The wattage capacity of the light bulb often does not match the size of the room, resulting in less than optimal lighting. Watt capacity or large power consumption sometimes also does not guarantee to be able to get good lighting. Therefore we need a decision support system that can calculate values ​​to be able to help users determine the desired light bulb properly and precisely according to needs. This decision support system implements the technique for order preference by simillarity to ideal solution (TOPSIS) method, which is a method that can give weighting and ranking for each criterion. With the technique for order preference by simillarity to ideal solution (TOPSIS) method, the author creates a system that is expected to later be able to assist decision making in the selection of electric light bulbs.
IMPLEMENTATION OF THE MOORA METHOD IN DETERMINING CANDIDATES FOR VILLAGE HEAD Fricles Ariwisanto Sianturi; Paska Marto Hasugian; Widia Putri; Ira Mayang Sari
INFOKUM Vol. 10 No. 03 (2022): August, Data Mining, Image Processing, and artificial intelligence
Publisher : Sean Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (253.388 KB)

Abstract

One of the community democratization parties is the election of the village head, the election of the village head is carried out at the village level directly to determine the village leader or village head. Decision support system is an interactive alternative system that can assist in decision making through the use of data and decision models to solve semi-structured and unstructured problems. This system was built by applying the MOORA (Multi-Objective Optimization on the basis of Ratio Analysis) method where the basic concept of the MOORA method is to find the weighted sum of the performance ratings for each alternative on all attributes. In the calculation of the MOORA method, only the one that produces the largest value will be selected as the best alternative. Calculations will be in accordance with this method if the selected alternative meets the predetermined criteria. By using a decision support system for the selection of village head candidates using the Multi-Objective Optimization on the basis of Ratio Analysis (MOORA) method, it helps community members to find out the ranking of village head candidates from the results of the weighted criteria that have been determined, thus providing additional information when making decisions. determine choices in the democratic party of the citizens of the village of Sitnggaling.
GENETIC ALGORITHM PERFORMANCE FOR ANALYSIS OF ELECTRONIC ORDER SCHEDULING Paska Marto Hasugian; Harpingka Sibarani
INFOKUM Vol. 10 No. 5 (2022): December, Computer and Communication
Publisher : Sean Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Order scheduling is the process of organizing and arranging orders to be fulfilled by a certain time and date. This can be done manually or through the use of scheduling software. The goal of order scheduling is to ensure that orders are filled in a timely and efficient manner, while also being mindful of any constraints or limitations that may affect the fulfillment process. This can include factors such as inventory levels, production capacity and delivery times. This study specifically discusses the scheduling of electronic goods orders. Scheduling electronic goods orders is the process of organizing and managing orders for electronic goods to be fulfilled at a certain time and date. This scheduling can be done using software or systems that can assist in the process of planning and managing orders for electronic goods. Several factors to consider in scheduling electronic goods orders are inventory level, production capacity, delivery time, and demand from consumers. In optimizing the arrangement of order distribution schedules using a genetic algorithm, four parameters are needed, including the number of destinations, the number of sales, order crossover probability and mutation.
Optimizing Data-Based Decision Making: Development And Implementation Of Decision Support System In Langkat Regency Government Sihotang, Jonhariono; Paska Marto Hasugian
INFOKUM Vol. 12 No. 04 (2024): Engineering, Computer and Communication, November 2024
Publisher : Sean Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58471/infokum.v12i04.2807

Abstract

Effective decision-making in local government is essential to improve the quality of public services and policies. The Langkat Regency Government faces challenges in managing data efficiently due to the limited integrated information system. This study aims to develop and implement a Decision Support System (DSS) to support data-based decision-making in local government. Using a mixed-method approach, data were collected through observation, interviews, questionnaires, and document analysis from related agencies. The developed DSS prototype integrates cloud computing and artificial intelligence to accelerate data analysis and generate policy recommendations. The results showed that DSS increased data processing time efficiency and decision accuracy by 40%. However, challenges such as resistance to change and limited infrastructure are still obstacles. Therefore, a strategy to increase human resource capacity and digitalization policies is needed to ensure the sustainability of DSS.
Co-Authors Agustinus Parmazatule Laia Al Hashim, Safa Ayoub Alex Rikki Amran Manalu Angelia M Manurung Anju Eliarsyam Lubis Annas Prasetio Arvind Roy Baehaqi Batubara, Muhammad Iqbal Betti Mastaria Br Sembiring Bobby Aris Sandy Bosker Sinaga Bosker Sinaga, Bosker Sinaga Br Ginting, Anirma Kandida Br Sembiring, Betti Mastaria Cinthya Agatha Sinaga Damianus Daha Devlin Iskandar Saragih Dewi Lasmiana Panjaitan Dharma Rajen Kartighaiyab Dharma Rajen Kartighaiyan Efendi, Syahril Emma Romasta Naulina Nainggolan Endang Utari Endra A.P Marpaung Fenius Halawa Ferdiansyah, Rahmat Fristi Riandari Fristy Riandari Giawa, Martinus Hanum, Rahmadiah Harefa, Ade May Luky Harpingka Sibarani Hasugian, Penda Sudarto Hengki Tamando Sihotang Herman Mawengkang Hidayati, Wenika Hutahaean, Harvei Desmon Hutahaean, Harvei Desmon Insan Taufik Ira Mayang Sari Jijon R. Sagala Jijon R. Sagala Jijon Raphita Sagala John Foster Marpaung Kristian Siregar Logaraj Logaraj Logaraj, Logaraj Logaraz Logaraz Lubis, Anju Eliarsyam Makmur Tarigan Manurung, Jonson Martinus Giawa Mathelinea, Devy Maya Theresia Br. Barus MIFTAHUL JANNAH Nababan, Adli Abdillah NASUTION, ATIKA AINI Ndruru, Risnamawati Nera Mayana Br.Tarigan Nico Setiawan Nurayni Sinabang Pandi Barita Nauli Simangunsung Penda Sudarto Hasugian Penda Sudarto Hasugian Poltak Sihombing Prawita Ardella R. Mahdalena Simanjorang Rahmat Ferdiansyah Riana Risnamawati Ndruru Ritha Zahara Tarigan Rizki Manullang Romanus Damanik Romauli Sianipar Sandy, Bobby Aris Sethu Ramen Sethu Ramen, Sethu Ramen Setiawan, Nico Siagian, Novriadi Antonius Sihotang, Jonhariono Sijabat, Petti Indrayati Simamora, Siska Simangunsong, Pandi Barita Nauli sinaga, lotar mateus Sinaga, Sony Bahagia Sinaga, Sony Bahagia Sinta Novianti, Sinta Sipayung, Sardo Sipayung, Sardo Pardingotan Siregar, Vanessa Sitanggang, Sarinah Situmorang, Caesar Juanda Theodorus Sri Wahyuni TONNI LIMBONG Uzitha Ram Vanessa Siregar Venentius Purba Vina Winda Sari Wenika Hidayati Widia Putri Yosapat Sembiring Yuda Perwira Yusi Tri Utari Panggabean