Tukino Paryono

This study is intended to evaluate and contrast the performance of varying combinations of embedding algorithms and weighting systems in measuring perception-based text similarity using the Cosine Similarity approach. Within a structured experiment design, a hybrid model referred to as HyEWCos (Hybrid Embedding and Weighting for Cosine Similarity) was built incorporating conventional embedding models (Word2Vec, FastText), transformer-based models (BERT, GPT), and statistical and linguistic word weighting schemes (TFIDF, BM25, POS-weighting, and N-weighting). The test results indicate that Word2Vec merged with the CBOW architecture and TFIDF weighting always returned the most reliable performance, with lowest error values (RMSE and MAE of 0.9868) and the highest rating correlation with expert judgment (Pearson’s, 0.524; Spearman’s, 0.543). These results show that contextually conditioned distributional representation approaches perform better in maintaining the semantic subtlety of short and subjective texts than transformer models that are not fine-tuned. This work is unique in terms of its evaluation framework because it integrates embedding and weighting approaches that have hitherto been examined mostly in separation. The main contribution of the study is the development of an experimental framework that serves as a foundation for building more stable and accurate text-based assessment systems. The research also proves the need for making decisions on representation methods based on the data type and domain and opens a door for continuing research in adaptive hybrid models and how their potential can be achieved through combining the best of various approaches.

Algoritma K-Means Clustering Untuk Segmentasi Produk Kosmetik Terlaris Berdasarkan Volume Penjualan

Poverty in Indonesia remains a major challenge, with significant levels of inequality between provinces. This study applies the K-Means clustering method to analyze poverty distribution patterns in 38 provinces in Indonesia, using data on the percentage of poor people from 2010 to 2024. With this approach, provinces are grouped into three main clusters: low, medium, and high, based on the average poverty rate. The low cluster includes areas with poverty rates below 10%, while the medium and high clusters indicate poverty levels that require more specific policies. The evaluation results show a silhouette score of 0.613, indicating that this grouping is quite good but can still be improved. This study offers important insights to support more targeted and effective policies, especially in achieving Sustainable Development Goal (SDG) 1: Eradicating Poverty.

The assessment document at the time of study program accreditation shows performance achievements that will have an impact on the development of the study program in the future. The description in the assessment document contains unstructured data, making it difficult to identify target indicators. Apart from that, the number of Indonesian-based assessment documents is quite large, and there has been no research on these assessment documents. Therefore, this research aims to classify and predict target indicator categories into 4 categories: deficient, enough, good, and very. Learning testing of the Indonesian language assessment sentence classification model using recurrent neural networks-long short-term memory (RNN-LSTM) using 5 layers and 3 parameters produces performance with an accuracy value of 94.24% and a loss of 10%. In the evaluation with the Adamax optimizer, it had a high level of accuracy, namely 79%, followed by stochastic gradient descent (SGD) of 78%. For the Adam optimizer, Adadelta, and root mean squared propagation (RMSProp) have an accuracy rate of 77%.

This training program aims to assist elementary school students in developing their fine motor skills by utilizing Paint software. Fine motor skills, which involve coordinating small muscles like the hands and fingers, are essential for children's growth and development. In this activity, Paint is used as a tool to enhance both students' creativity and their motor abilities. Students engage in activities such as drawing, coloring, and pattern-making, which improve hand-eye coordination and establish fundamental artistic skills. The outcomes of this training reveal significant improvements in students' fine motor skills as well as their proficiency in using technology for creative purposes.

Public sector digital transformation requires a deep understanding of the digital readiness of each administrative region. The Electronic Government System (EGIS) Index is used by the Government of Indonesia as a measuring tool to assess the digital maturity of government agencies. This study aims to cluster districts/cities in West Java Province based on their 2023 EGIS scores to identify hidden patterns of digital readiness. Three unsupervised learning algorithms—K-Means, DBSCAN, and Agglomerative Clustering—are used to explore data-driven regional segmentation. The analyzed dataset includes 27 administrative regions and a number of numerical features related to the EGIS dimensions. The results show that each method is able to form clusters that reflect variations in digital readiness, with DBSCAN producing the most detailed segmentation and being able to detect outliers. Agglomerative Clustering shows good hierarchical separation, while K-Means provides a fairly representative general division. This study provides an analytical basis for contextual and targeted cluster-based policy making in developing regional digital transformation.

Puskesmas is a location for top-notch volunteer work that benefits the village and city governments alike. Therefore, patient feedback regarding the kinds of services offered by the community health center is required in an attempt to improve the quality service performance. Patient opinions can be expressed through reviews or opinions about the quality of patient care on social media sites like Facebook, Instagram, Twitter, WhatsApp, and Instagram. On the other hand, thoughts shared on social media are lengthy, unstructured texts. This complicates text analysis and makes it impossible to compare the caliber of services offered by Puskesmas managers. Furthermore, a number of Community Health Centers lack websites that allow users to rank Community Health Centers according to user interest and visual appeal and efficiency in operations. Thus, the purpose of this study is to classify and present sentiment analysis from Twitter about community health centers' health services. The scope focuses on five factors: administrative services, finances, mechanisms, health worker friendliness and skills, and administrative services. The LSTM word embedding model and the adadelta and adamax optimizers are used in word embedding for text mining. A confusion matrix was used to evaluate the developed model's degree of accuracy in categorizing and forecasting patient reviews. Results from the LSTM and Adamax models with a precision level of 76%, Recall 69% and Accuracy 71%. The results of this research show that the LSTM method and Adamax optimizer can classify and predict public opinion data about Puskesmas services via Twitter quite well. A high level of accuracy is very important to ensure that community opinions can be properly identified by the model, so that it can support the decision-making process in improving the type of Puskesmas services. To improve the model, further studies can be conducted on how to select parameters, select features, and create a quality dataset.

Plates continues to increase. This research aimed to detect vehicle license plates using ESP32CAM and utilize photo text reading using Google Lens, which can be used online to retrieve numeric characters. The method approach was to connect Wifi connectivity to the ESP32CAM, which had been programmed to detect vehicle plates. Vehicle plates that have been detected and recognized were inputted into Google Lens to capture the resulting text from the ESP32CAM camera recording. The results of this study were that for 70 seconds, ten plate samples were obtained, which were 100% perfect in reading license plates on Google Lens, namely six plates and two plates read 90%, one plate read 60%, and one plate read 0%. The research conclusions obtained were ten samples, six samples with perfect readings, and one error sample because of the white plate color. Thus, the main objective was to obtain the results of the vehicle plate detection and retrieve the text from the recording results

A credit card is one of the legal payment media owned by a bank in making a payment transaction within the agreed timeframe. In particular, credit services are provided by institutions or bodies that have the authority to distribute funds in the form of financial assistance to individuals and groups. However, in practice there are bound to be obstacles, especially during payback periods that often occur, such as when a customer wants to submit a Repeat Order or apply for funds again. Obstacles that are usually encountered in the process of granting credit are substandard credit and bad credit payments. Before PT Esta Dana Ventura wants to decide to approve applications for re-granting credit cards from prospective repeat order customers, a classification of assessment criteria is needed to determine the feasibility of granting credit to prospective repeat order customers. This study made the decision to use Data mining clustering classification with Rapidminer tools as a tool to obtain accurate results by processing data using the K-Means clustering method to help PT. Esta Dana Ventura in analyzing potential non-performing loans. By comparing survey data for Repeat Order candidates with previous credit granting data and classifying them in the form of bad or non-bad credit classifications.From the results of research using the k-means method it can produce grouping data into 3 criteria, namely (C0) 69 data with current customers, (C1) 3 data with very current customers, and (C2) 52 data with Bad customers..

The technology of the metaverse has been a topic of discussion in the tech sector recently, and the implementation of the metaverse in the future is one of the most anticipated things for internet users around the world, allowing users to become the avatar they want in the virtual universe. To see how someone approaches metaverse technology, we can analyze it by looking at various opinions that emerge from the sentiment of internet users, one of which is from the social media Twitter. This study aims to analyze the sentiment of Twitter users' opinions, which is one of the most popular internet social media platforms. The researcher collects tweet data using the snscrape library in the Python programming language, a tool for mass data collection from Twitter. Then, the analysis of the sentiment of Twitter user opinions is carried out using the Lexicon Based method, which is a method that uses a dictionary of words to categorize the sentiment of a text. The results of this study show that the Lexicon Based method can classify the sentiment of Twitter user opinions related to metaverse technology with an accuracy value of 84%. This shows that the results of sentiment analysis using the Lexicon Based method have a good level of accuracy in performance in analyzing the sentiment of Twitter user opinions related to metaverse technology.