Publication Search

This community service program is designed to help MSMEs in Kedung Waringin Village utilize TikTok as a creative marketing medium. A total of 22 business actors participated in the training which included theory and practice, including understanding the TikTok algorithm, creating video content, and using features such as hashtags and TikTok Shop. The results of the training showed a significant increase. Before the training, 60% of participants did not understand digital marketing, but this figure increased to 85% after the training. In addition, 90% of participants felt that the training was relevant to their business needs, and 60% of them recorded an increase in sales of up to 30% after implementing the strategies taught. This activity provides direct benefits by improving participants' skills in digital marketing and building their confidence to use technology independently. This program shows the success of a practice-based approach and can be a model for MSME empowerment that can be applied in other areas

In today’s era, social media has become a driving force for increasing digital entrepreneurship. Businesses are utilizing social media sites such as Instagram, TikTok, LinkedIn, or even Facebook to brand their companies or products and interact with clients. This is great news for businesses, especially SMEs, to have low-cost access to key markets worldwide. One evident trend is the emergence of social commerce – business-to-consumer commerce without intermediaries, exclusive of other e-commerce models. However, the adoption of social media in digital entrepreneurship comes with several challenges, such as changes in algorithms that can affect content visibility and risks related to data security and user privacy. Nevertheless, social media remains useful in terms of analytics to support strategic decisions. This study shows the value of social media for entrepreneurship and technologies that help improve content personalization and consumer behavior analysis, such as artificial intelligence and big data. This study attempts to fill the gap in the literature by looking at the differences in the outcomes of social media use in developing and developed countries and the outcomes of new technologies on digital business ventures.

This research explores increasing the efficiency of network resource use through load balancing techniques. Load balancing distributes workload evenly among servers to optimize resource usage, maximize throughput, and minimize response time. We compare various load balancing algorithms, such as round-robin, least connections, and least response time, in various network scenarios. Experimental results show that proper load balancing techniques can reduce response time by up to 30% and increase resource usage efficiency by up to 40%. This study highlights the importance of selecting appropriate algorithms based on network traffic and workload characteristics. Implementing effective load balancing strategies can improve the quality of network services and ensure even distribution of workloads, providing practical guidance for network administrators to optimize network performance and efficiency.

Computational thinking is the ability to solve problems and design systems using concepts and techniques generally associated with computers and computer programming. The aim of this research is to conduct a literature review, namely to determine the application of computational thinking in mathematics learning. The technique used is the Systematic Literature Review (SLR) strategy. SLR is a research method that aims to identify, study and interpret data in journals systematically with specified stages. The conclusion obtained from the research results is that the application of computational thinking in mathematics learning can be done using four foundations of computational thinking, namely: (1) Decomposition, namely the problem is divided into small parts, (2) Pattern recognition, namely the process of identifying patterns or sequences of problems, (3) Abstraction, namely consideration of the important parts of a problem, and (4) Algorithm, namely a series of instructions for solving the problem. Examples of its application include the use of interactive mathematics software, mathematical simulations, problem-based learning with technology, adaptive learning and the use of educational games.  

The rapid advancement of quantum computing has significantly impacted data security, as classical cryptographic algorithms such as RSA and ECC are increasingly vulnerable to quantum attacks. This study aims to evaluate the performance of classical and post-quantum cryptographic algorithms in a quantum simulation environment, focusing on stability, efficiency, and computational time. The research method employed experimental simulations using Qiskit, where cryptographic algorithms were modeled into quantum circuits and tested across varying qubit sizes of 128, 256, 512, and 1024. The simulation results indicate that classical algorithms face substantial limitations, with exponentially increasing computational time and drastically reduced stability beyond 512 qubits. In contrast, post-quantum algorithms demonstrated superior performance, maintaining high stability up to 1024 qubits, achieving greater quantum efficiency, and showing resilience against quantum attacks such as Shor’s and Grover’s algorithms. These findings highlight the urgent need to transition toward post-quantum cryptography as a more adaptive and reliable approach to safeguarding data in the quantum era. Although post-quantum algorithms still face certain challenges, such as larger key sizes and slightly higher computational costs at smaller scales, their overall benefits are far more significant in ensuring sustainable information security. Therefore, adopting post-quantum cryptography represents a strategic step that must be prioritized to address the evolving risks posed by quantum computing technologies.

Abstract.  The purpose of this research is to identify the difficulties experienced by students in understanding the concepts, principles, and procedures in problems related to the material of lines and angles. This research is a qualitative study with a descriptive method. The subjects of this study were students of class VIII-1 at SMP Negeri 29 Medan. Data collection techniques included tests, interviews, and documentation. All students of class VIII-1 took a test on the understanding of concepts, principles, and procedures. Afterward, the test results were analyzed and categorized into three groups: high, medium, and low ability. Subsequently, two students from each ability group were selected as interview subjects. The results of this research indicate that the most common conceptual difficulty experienced by students is classifying objects according to their concepts in determining the type of angle according to its measurement. The most common principle difficulty experienced by students is connecting mathematical concepts in solving problems in the material of the relationship between angles formed by two parallel lines intersected by a transversal line. Meanwhile, the procedural difficulty experienced by students is that all students still make mistakes in explaining, planning, and applying the correct steps or algorithms in mathematical problems, both in the material of the relationship between lines, determining angles, and the relationship between angles formed by two parallel lines intersected by a transversal.  

With the rapid development of information technology, Tiktokshop has become an increasingly popular e-commerce platform in Indonesia. However, selling on Tiktokshop does not always guarantee success, especially for small and medium enterprises (MSMEs) that focus on selling special products such as CustomKids. This research aims to determine the impact of Tiktokshop's inactivity on sales of MSMEs, especially the custom children's industry in Indonesia. This research analyzes various factors that might influence Tiktokshop's inactivity, such as lack of presence, lack of user interaction and algorithm changes. We also assessed the direct impact of the suspension on sales of CustomKids products, taking into account variables such as decreased sales, decreased market share, and the psychological impact on MSME owners. The results of this research will provide insight to MSME owners, the government and stakeholders to help them develop strategies to advance Tiktokshop activities and increase the competitiveness of MSMEs, with a special focus on CustomKids products. It is hoped that the implications of this research finding will help in decision making to improve the performance of MSMEs in the digital era.

This research allows an automated system for detecting classified means of transportation in Medan City traffic using the YOLOv8 algorithm. The YOLOv8 algorithm is used to detect transportation objects with accuracy that is many times better than other object detection algorithms and with good accuracy after training with various data sets. The use of this algorithm provides an effective solution for handling congestion in the form of increasing the number of vehicles and less orderly traffic users in the city of Medan. The placement of each transportation object in the image to be tested by the system has an influence on the shape accuracy of the object detection results by the algorithm.

Quantum computing has emerged as a revolutionary paradigm, holding immense potential to solve complex problems that classical computing struggles to address. This study explores the application of quantum computing in cryptography, with a specific focus on two major quantum algorithms: Shor’s algorithm for large number factorization and Grover’s algorithm for unstructured database searching. The main objective of this research is to compare the performance of these quantum algorithms with classical cryptographic methods in terms of computational efficiency and time. Shor’s algorithm, which can factorize large numbers in polynomial time, presents a significant threat to the security of public-key cryptosystems such as RSA, which rely on the difficulty of factoring large numbers. On the other hand, Grover’s algorithm offers a quadratic speedup for searching unstructured databases, making it highly relevant for symmetric key cryptography systems like AES. In this study, simulations of both algorithms were conducted using quantum simulators to assess their speed and effectiveness in solving cryptographic challenges. The results demonstrate that quantum algorithms significantly reduce the computation time compared to classical methods, with Shor’s algorithm efficiently solving factorization problems and Grover’s algorithm accelerating key searching processes. However, while these quantum algorithms show promise in improving cryptographic systems, the implementation of large-scale quantum computers remains a challenge. This research highlights the potential of quantum computing to revolutionize data security and underscores the need for further development in quantum algorithms and the transition to quantum-resistant cryptographic systems to safeguard against the threat posed by quantum computers.

This study aims to analyze the communication of marine mammals, especially whales and dolphins, through a bioacoustic approach combined with computational science as an effort to support conservation in the Tropical Ocean region. The focus of the location is on the Banda Sea, the Seram Sea, and the tropical Pacific region which are important migration routes for marine mammals. Data were obtained from underwater sound recordings using hydrophones, accompanied by visual observations to validate the behavior and existence of species. The analysis is carried out through several stages, including signal pre-processing with noise filtering and sound segmentation, spectral analysis using Fast Fourier Transform (FFT), as well as the creation of a spectrogram to visualize vocalization patterns. Machine learning algorithms such as Support Vector Machine (SVM) are used to classify interspecies voices, while deep learning approaches are applied to identify more complex communication patterns, including dialect variations. The results showed that whales produced low-frequency vocalizations (20–200 Hz) for long-distance communication, while dolphins used high-frequency clicks and whistles (5–20 kHz) for echolocation and social interaction. The integration of bioacoustics and artificial intelligence improves the accuracy of sound classification by more than 90%. These findings confirm the effectiveness of computational-based non-invasive methods in monitoring the presence and behavior of marine mammals and provide a scientific basis for sustainable conservation.

The FishGo application is a navigation-based android application using remote sensing sensor technology that is processed with a certain algorithm so that it can provide information on potential fish areas and the best route for traditional fishermen to catch fish. The Patriot Tool (Detecting Fish Catching Areas Using the Internet of Things System) is a development of an application in the form of a fish finder tool to more accurately detect underwater fish biomass. The research objective is to determine the success of patriot service innovation through the FishGo application as an effort to improve the welfare of fishermen in Badung Regency. The research method is descriptive qualitative with data collection techniques in the form of observation, interviews and documentation. The research analysis uses the theory of innovation success factors put forward by Bugge, et al which consists of 6 indicators: Governance and innovation, Sources of ideas for innovation, Innovation culture, Capabilities and tools, Objective, outcomes, drivers and trouble, and Collecting innovation data for single innovation. The results of this study are that the implementation of innovation has achieved good results on the source of innovation ideas including internal and external, a culture of innovation by showing good changes to fishermen, cooperation in developing innovation and socialization activities that are intensively carried out to fishermen, the results felt by fishermen experience catch increase. However, innovation governance is still not optimal, there are no basic regulations such as regional regulations and SOPs governing innovation, Human Resources capabilities still need to be added to the fields of electronics, IT and mapping, goals that have not been fully achieved from those that have been set. and there are still external and internal obstacles in its implementation.

Quantum-Inspired Algorithms (QIAs) combine principles of quantum computing with classical evolutionary strategies to address complex optimization problems. This research explores the potential of QIAs in improving optimization processes, particularly in combinatorial and multi-objective optimization scenarios. The study focuses on the application of Quantum-Inspired Genetic Algorithms (QIGAs) and Quantum-Inspired Evolutionary Algorithms (QIEAs), assessing their effectiveness in solving classical problems like the Traveling Salesman Problem (TSP) and Minimum Spanning Tree (MST). Through computational simulations, the research compares the time convergence and solution accuracy of QIAs against traditional classical algorithms. The findings demonstrate that QIAs achieve faster convergence rates and higher-quality solutions, with accuracy levels reaching 98-99% of the global optimal solutions, while significantly reducing computational time. These results underline the advantages of QIAs in solving large and complex optimization problems, making them a promising alternative to traditional algorithms. Additionally, QIAs excel in avoiding local minima, a common pitfall of classical methods, due to their ability to explore the solution space more efficiently through quantum principles like superposition and interference. The implications of this study suggest that QIAs can be a valuable tool for tackling real-world optimization challenges, with potential applications in fields such as finance, logistics, telecommunications, and energy management. The research also indicates the necessity for further improvements in quantum-inspired algorithms' scalability and hardware integration, particularly for larger, more intricate optimization problems, to fully realize their potential in practical industrial applications.

The Toba Batak tribe has a distinctive fabric known as ulos. Toba Batak ulos have types depending on their uses. But in the modern era, especially among urban communities, very few people know the types and uses. Motivated by the success of Convolutional Neural Network (CNN) algorithm in image classification, this study will conduct a learning-based approach to classify 5 types of Toba Batak ulos (Ragi Hidup, Ragi Hotang, Mangiring, Sadum, and Sibolang). The process starts from data collection, data analysis, model building, model training, and confusion matrix. The dataset used is 1000 images with 80% training data, 10% valid data, and 10% test data. Convolution, maxpooling, dropout, flatten, and fully connected are the 5 layers forming the CNN model. The optimizer used is Adam with a learning rate of 0.001. The model generated in this study can detect Toba Batak ulos images at an accuracy rate of 94.00%.    

Travelling Salesman Problem (TSP) is a problem that is often encountered by a salesman who must travel exactly once to all consumers in a route and will return to the starting point of departure. Algorithm Genetic Algorithm is one way to find heuristic solutions based on the evolutionary ideas of natural selection and genetics. The aim is to find the optimal route for the distribution of bottled water products produced by PT. Mual Natio Maju Bersama. To find a solution, the chromosomes processed by the genetic algorithm are represented through the stages in the Genetic Algorithm individual initialization, fitness value, linear fitness ranking, roulette whell selection, crossover, and mutation. In order to achieve the optimum solution, namely The best path obtained is PT Mual Tio Maju Bersama –BUMDES Sait ni Huta - UD. Alvaro - UD. Lancelhot – UD. Alris – UD. Jamel – Toko Kelontong SRC Resi 2 – Toko Notra – UD. B Siringoringo – Toko Dahlia Siahaan – UD. Purba – UD. Cahaya – UD. Hutapea – UD. Gabe – UD. Setia II – UD. Larisma II – UD. Antoni – UD. Bona Siahaan – UD. Sederhana – Toko Manalu – UD. Setia I – Toko Ferdinan – UD. Alboy – Wisma Daun Mas – UD. Top Jaya – UD. Mega Silaban – BUMDES Silaitlait – UD. Rika – UD. Panamot – Piltik Coffee and Homestay Bandar Udara Silangit – UD. Rolas Boy – UD. Salamat Karya – UD. Simpang Jaya – UD. Lambok - Piltik Coffee and Homestay Siborongborong – UD. Bahagia – UD. Marlinca – UD. Heri Joel Pasaribu – UD. Ebenezer – UD. Mawar – UD. A Saudara – UD. SP Perdana – PDAM Mual Na Tio – UD. Rokkap - PT Mual Tio Maju Bersama. The best path length is 125.2700 cartesian units and the best fitness value is 0.008000.

The purpose of this research to revisits methods that are more effective in nonlinear Optimization with single variable polynomial functions at high degrees. Models with linear objective functions and constraint functions are Polynomials of third, fourth and fifth degree reconstructed into subproblems that are easier to solve, namely quadratic programs, using bilinear Auxiliary Functions and solved by MATLAB simulations. The method used is the development of Tawarmalani & Sahinidis' research regarding relaxation with Auxiliary Functions. Examples of nonlinear Optimization with polynomial functions are also given to illustrate the implementation of this algorithm. The results of the research show that the application of the development reconstruction method produces a global solution that is no better than the solution to the original problem so that it is not an effective alternative method to use.

With the increasing development of technology the more variety of books circulating on the internet. As is the recommendation system on online book sites that provide books relevantly and as needed with one's preferences. One alternative is GoodReads, a social networking site that specializes in cataloging books and users can share reading book recommendations with each other by rating, reviewing, and commenting. As a large book recommendation site, it has a lot of data that can be processed by applying machine learning methods, but still not known as the most accurate model. By using the right model, we can provide more accurate recommendations. Therefore, this study will analyze the data obtained from the www.kaggle.com namely the goodreads-books dataset. This study proposed a data mining classification model to get the best model in recommending books on GoodReads. The algorithms used are Decision Tree, K-Nearest Neighbor, Naïve Bayes, Random Forest, and Support Vector Classifier, then for model evaluation using accuracy, precision, recall, f1-score, confusion matrix, AUC, and Mean Error Absolute. The test results of several classification algorithms found that Decision Tree has the highest accuracy among the methods presented by 99.95%, precision by 100%, recall by 96%, f1-score of 98% with MAE of 0.05 and AUC of 99.96%. This is proof that decision tree algorithms can be used as book recommendations based on book categories on GoodReads.

Abstract In this study that was used as the object of research in classifying red wine based on the quality influenced by each red wine or red wine based on the content of each type of wine, from each attribute containing the composition in the wine seen which attributes most affect the quality of red wine, so that it will be known ingridents that can improve the quality of the wine, in this study was carried out by the application of Machine learning by comparing three algorithms of mining data that is , Decission Tree, Random Forest and Support Vector Machine (SVM), from the results of research that has been done by comparing the three algorithms, Random Forest produced the best accuracy among other algorithms that have been tested. Random Forest with accuracy results of 0.7468 makes this algorithm best used to classify the quality of red wine. And in the second order Decission Tree with accuracy results of 0.7031, while Support Vector Machine (SVM) get an accuracy result of 0.65. So in the research that has been done to classify the quality of red wine based on its composition Random Forest becomes the best algorithm to use..