http://repository.president.ac.id/xmlui/handle/123456789/11834 Tue, 21 Apr 2026 09:43:43 GMT 2026-04-21T09:43:43Z http://repository.president.ac.id/xmlui/handle/123456789/11940 BOOK RECOMMENDATION SYSTEM WEB BASED USING K-NEAREST NEIGHBORS ALGORITHM AND COLLABORATIVE FILTERING Nabilla, Raykhana A recommendation system is a technology that can facilitate internet users in finding things. The recommendations given can be in the form of information, goods, and others. This system has been widely used in various platforms including social media, e-commerce, and many others. In this paper, the recommendation system made is about the recommendation system on books. There are many types of books currently, books about education, novels or fiction stories, comics, biographies, self-development and others. However, because of the many choices of books that can be read, readers will usually find it difficult to choose a book. To overcome this, readers will usually look for a recommendation through friends or through the internet. In the recommendation system there are several methods that can be used, namely content-based filtering, collaborative filtering, and hybrid filtering. Hybrid recommendation is a combination of two or more recommendation systems. In this paper the method used is collaborative filtering, the way it works isto add up the ratings or choices of a product, find user profiles by looking at the rating history given by users, then generate new recommendations based on comparisons between user patterns. The algorithm used in this book recommendation system is the K-Nearest Neighbors algorithm. The K-Nearest Neighbors algorithm is one algorithm that can provide recommendations with good accuracy. This system helps to save readers time in searching for books so that readers can make the right decision about the book to be read next, the recommendations given by the system to users are expected to be in accordance with the preferences or interests of each reader who uses this application. Sun, 01 Jan 2023 00:00:00 GMT http://repository.president.ac.id/xmlui/handle/123456789/11940 2023-01-01T00:00:00Z http://repository.president.ac.id/xmlui/handle/123456789/11939 AQUA ALERT : REAL-TIME OBJECT DETECTION APPLICATION FOR DAILY REMINDER TO DRINK WATER USING TENSORFLOW WITH CONVOLUTIONAL NEURAL NETWORKS Pratama, Muhammad Yoga With the rise of deep learning and the improvement of computer vision, object recognition has become a popular study topic. This is mostly because of convolutional neural networks (CNNs). By using a lot of labeled data, CNNs have shown that they are good at finding and predicting things in pictures. This project uses this skill to solve a common health mistake made by office workers: not drinking enough water every day. The Ministry of Health stresses how important it is to drink enough water for your body weight. Still, many workers forget about this health tip because they are too busy with their daily work. The main goal of this project is to set up a real-time object recognition system that will tell these people to drink water at the right time. Built on the flexible TensorFlow framework, the system is designed to be as efficient as possible, so it can work even on machines with less powerful hardware. The only thing you need to use it is a webcam, which makes it a useful and easy-to-use option. The project was inspired by real-life experiences during an internship. It also takes into account the fact that workers often don't drink enough water because of their busy work schedules, even though they work in air-conditioned settings that make thirst worse. This mistake can lead to health problems like dehydration, heart problems, and stomach problems. The suggested system is meant to fit right into their work surroundings, reminding them to drink water regularly and helping them live a better life at work. By combining cutting-edge technology with a basic health need, this program aims to get people to drink more water, which will improve their health and comfort at work in the long run. Sun, 01 Jan 2023 00:00:00 GMT http://repository.president.ac.id/xmlui/handle/123456789/11939 2023-01-01T00:00:00Z http://repository.president.ac.id/xmlui/handle/123456789/11938 ARTIFICIAL INTELLIGENCE CHATBOT APPLICATION FOR OPTIMIZING LEGAL SERVICE Pahlevi, Reza Febran In the field of law, especially legacy, there are many pasal and laws that are listed. And if someone is related to the law and needs certain pasal then someone must find and validate it accurately with their needs combines a technological advance and the world of law and legacy that is integrated with chatbots and ai and can create an interface with its users and lighten the burden for those who need it.Due to the author's concern about this issue, the author developed a system to help the who need for finding some pasal and legacy using web based and AI bot. For this project the author decided to make a web based project can filter pasal with integrated Artificial Intelligence then applied to the chatbot to make it easier for someone to qualify pasal and laws that are needed by users who are looking for pasal or legislated in need. With several features on the Web, the author provides the sorting of pasal needed - Chatbot for looking for the required article components and then calibrating it with technology "speech" the user can search for a component just by speaking then the chatbot will catch what the user means. Sun, 01 Jan 2023 00:00:00 GMT http://repository.president.ac.id/xmlui/handle/123456789/11938 2023-01-01T00:00:00Z http://repository.president.ac.id/xmlui/handle/123456789/11937 FACE RECOGNITION ATTENDANCE SYSTEM WITH AUTOMATION Fransisko, Sinaga, Yohanes This thesis describes a face recognition attendance system that utilizes the Local Binary Patterns Histogram algorithm for the face recognition and Flask framework for the backend server and the automation. The primary objective is to create a system that is more efficient for the education sector which now the manual attendance system is still widely used in this sector. The automation is to send an attendance report to the person in charge, which could be the lecturer, teacher, or admin person. The tool used in this thesis for the system is Python for the programming language as it is commonly used in machine learning. XAMPP to handle the database, Flask framework to handle the backend server, and connect with the database. The benefit of this thesis is enabling the education sector to implement the attendance system efficiently with less cost as the development of the system used open-source frameworks and also the improvement in the attendance system. The result of this thesis is a face recognition system is that it works as expected with the features listed in the system overview as the parameter. Sun, 01 Jan 2023 00:00:00 GMT http://repository.president.ac.id/xmlui/handle/123456789/11937 2023-01-01T00:00:00Z