Journal of Computer and Creative Technology

Developing System of Students' Home-Visiting by Route Analysis Technique and Cloud Maps

Nopporn Krairak, Wudhijaya Philuek — Wed, 16 Oct 2024 00:00:00 +0700

This study aims to develop a web application for students' home-visiting by route analysis technique and cloud maps in the data of primary education students of the schools in Nakhon Sawan Province. Consideration of key data characteristics, such as learning interest, academic performances, and genders, from data testing with Decision Tree and Neural Network. The results found that 1) the web application is operated by analyzing to design a use case diagram, the first records basic students' data, the second analyzes the recorded data using a Decision Tree, and a Neural Network to analyze the data and predict which student's home the teacher should visit in the data set and summary of students’ home- visiting results, the third data export to pdf format, 2) digital platform design, 3) development, and installation, 4) testing, 5) actual installation, 6) maintenance.

Effects of Using Applications to Thai Phonetic Recognition According to the Reading Test of Grade 1 Students

Somphon Passon, Wudhijaya Philuek — Tue, 15 Oct 2024 00:00:00 +0700

This research aims to study the effects of using applications to Thai phonetic recognition according to the reading test of grade 1 students. The data used in the study were 800 voice readings with 40 students by using the sound feature extraction including Spectrogram and Mel Frequency Cepstral Coefficient techniques in conjunction with the use of Neural Networks to train data and test models for application development. The study found that pronunciation training and similarity comparison from sound signals after using an application that can analyze the pronunciation of 800 sounds, and Thai pronunciation of 160 words in the reading test has an accuracy of more than 50%, which is 93.75%, and less than 50%, which is 6.25%, after using the application to detect the pronunciation, the accuracy is at a very good level.

Comparing a Thai Words Segmentation Methods in the LST20 Dataset

Krittapol Damrongkamoltip, Khatcha Ruenlek, Wasit Limprasert, Prachya Boonkwan — Mon, 12 Aug 2024 00:00:00 +0700

In this era of globalization where information is widely available, organizations are increasingly placing importance on using information to enhance their business. Although data is easily available, there are still challenges in natural language processing tasks, especially, the division of Thai words that lacks clarity of word boundaries, etc. This makes it difficult to identify the word groups in a sentence appropriately. Therefore, this study focuses on evaluating the performance of the word segmentation method including the Dictionary use and learning from data using evaluation of word segmentation in six techniques are important goals for the verification of the literal level accuracy and processing time of each method and technique, by the LST20 dataset contains 3,745 documents and covers 15 news categories in results show a more efficient way to learn from data.

Land Price Prediction with Machine Learning in Mueang Khon Kaen District

Thu, 15 Aug 2024 00:00:00 +0700

This research aims to develop a model of land price assessment and study of factors influencing land prices with machine learning used as a guideline for determining the estimated price to be close to the actual purchase price in Mueang Khon Kaen District from land price information traded on websites in enforcement department of 193 locations, and land price assessment information in land department of 1,500 locations in this study the factors involved include appraisal value, property type, land size, distance, and the average appraisal value of five nearby plots of land. The models used for analysis are Regression Tree, Random Forest, Gradient Boosted Trees, and Linear Regression. Which, the land price assessment from the case study by model measurement in MAE, RMSE, R-squared, Grid Search, and Cross-validation to selected model parameters and evaluate their performance. The results found that the model with the best predictive performance is Gradient Boosted Trees in R-squared at the highest of 0.80, MAE, and RMSE at the lowest of 7929.40, and 15281.33, respectively. Feature Importance in the locations with the most influence on prediction, followed by area size, average appraised value from five nearby locations, and property type.

Activities Calendar System in Faculty of Interdisciplinary, Khon Kaen University

Thu, 10 Oct 2024 00:00:00 +0700

Public relations operations, news, and various activities in the Faculty of Interdisciplinary, Khon Kaen University in the past, there was still a lack of a main channel for publicizing the faculty's and department's central activities, which resulted in a small number of participants and an ongoing impact on the organization of the event. This research aims to design and develop of activities calendar system in the form of a web application, by applying the DevOps software development approach, consisting of steps for studying and analyzing problems, designing and developing systems, system testing, system delivery, and satisfaction evaluation of use in student and teacher development groups. However, the system includes the event calendar function, and event details, advance notification, announcement management, adding, deleting, editing events, event registration, sharing event links, attendee reporting, attendee statistics, and certificate issuance. Satisfaction with the use of the display screen was at a high level, and usability was at the highest level.