2022 International Conference on Information Technology Education (ICITE 2022) Sentiment Analysis-based Web Application for Digital Tourism Marketing Divine Aguilar and Cecilia Mercado 08:20-08:40 Abstract: The COVID-19 pandemic greatly affected tourism. 6845 Concerning this, digital technologies are integral in helping countries overcome and restart tourism activities in the new normal. This paved way for the need for this research, which created a blueprint that enables crowdsourcing sentiments from tourists and processes the sentiments. In order to start building the model for the application, data collection process was undertaken first. Under the data collection process, over 100,000 tweets were gathered from Tweeter wherein after initial pre-processing only 24,980 were retained. The project was created using Tensorflow and Keras to perform sentiment analysis into positive or negative. To come up with the best number of epochs, the study exhausted over 100 epochs for the machine-learned model. The dataset was divided into two datasets wherein 80% of the dataset was used for training the model and 20% was used for testing the model. The findings of the study shows that around 15-20 epochs the training and validation accuracy were at par with each other having around 85-90% accuracy. The result also shows that the higher the epochs the more overfitting happens wherein the training results in higher accuracy as the epochs increases and the validation results in a fluctuating pattern. Guided by the RAD Model, this study created a sentiment-based web application by gathering and identifying user requirements and requirements for building the machine-learned model, designing the data model and the user interface design and lastly integrating a machine-learned model into a web application. 08:40-9:00 Eye-M-Safe: A Computer Vision-based Overcrowding Detection 8309 and Prevention System for Covid-19 Risk Reduction in Public Establishments John Therende D. Calma, Mark Allen A. Dela Torre and Christian F. Quijano Adamson University, Metro Manila, Philippines Abstract: In this study, the researchers developed a bipartite system that encompasses the creation of a desktop application 50
2022 International Conference on Information Technology Education (ICITE 2022) that sought to prevent and control overcrowding without the need for human resources by utilizing computer vision in the 9:00-9:20 automation of detection and prevention of overcrowding in 7705 public establishments, specifically through the implementation of convolutional neural networks (CNN) for object classification and detection, and the development of an automated overcrowding assessment and alert system for overcrowding assessment and prevention; furthermore, the bipartite system also encompasses the creation of an android application that seeks to provide convenient and accessible indicators for the general public to use as a means of a decision-support system, specifically to whether a nearby EYE-M-SAFE-enabled establishment is safe or not, based on their overcrowding safety level (OSL) and occupancy level, thereby assisting them in avoiding the risk of getting infected due to overcrowding in public establishments, and ultimately assisting in the COVID-19 risk reduction in public establishments. Pest Classification and Prediction: Analyzing the Impact of Weather to Pest Occurrence Through Machine Learning Evan C. Sumido - College of Information and Communications Technology, West Visayas State University, Luna St. Lapaz, Iloilo City, Philippines Larmie S. Feliscuzo and Chris Jordan G. Aliac - College of Computer Studies, Cebu Institute of Technology - University, Natalio B. Bacalso Ave., Cebu City, Philippines Abstract: In the Philippines, pest infestation is the most common problem that affects rice production and the profits of our farmers. Anticipation may mitigate possible damages that a pest can bring. This study aims to classify and predict pests with the use of machine learning techniques. This will also attempt to identify the contributing variables that has an impact in terms of its accuracy. Prediction is based on the values of different weather parameters during infestation. Datasets containing weather variables was taken from an Automated Weather Station located on the town where the recorded infestation happened. Stem borer and brown plant hopper are the most common pests that infected the area and were selected for the study. Machine learning techniques such as Random Forest, Naïve Bayes, XGboost and CART were utilized and compared. 51
2022 International Conference on Information Technology Education (ICITE 2022) It showed that Random Forest and XGBoost algorithms performed excellently in terms of classification and prediction. 9:20-9:40 For most machine learning techniques used, solar radiation and 7850 dew point found to have the higher influence affecting the accuracy of the model in performing prediction and classification. For future studies, results may be used and served as basis in developing an Integrated Pest Management System and Decision Support System. Penetration Testing Engagement Web Application Tool Marlon Tayag and Francisco Napalit A Real-Time Web-Based Street-Level Mapping System for Traffic Monitoring D arwin Mangca, Larmie Feliscuzo and Elmer Maravillas Cebu Institute of Technology University 9:40-10:00 Abstract: This paper focuses on a web-based street-level 8642 mapping system that could process real-time data coming from the streets and can be accessed online by the public. The advantage of using this digital tool is that it can lessen traffic congestion by allowing the public to know which areas are congested and avoid them. It can also help the top management to understand the current traffic condition in a city view level and do some appropriate actions to lessen the congestion. The method used is rapid application development which is suitable for system development. The simulation results showed that the system was able to be viewed by the public and can add classes per street. The system is scalable and convenient when adding a new data source coming from an IP camera. Predictive Analytics of Children Vaccine Supplies in Negros Occidental Philippines Lynol I. Ibarra, May S. Cuaycong and Mae B. Lodana 10:00-10:20 Abstract: Data modeling gained popularity and practical 8162 usability as more organizations shifted from manual to digital transactions. Most of the time, data models were integrated into a Decision Support System (DSS) to assist in the organizations' decision-making and provide insights based on the available data. 52
2022 International Conference on Information Technology Education (ICITE 2022) Negros Occidental, the Health Centers are on the front lines of the child immunization program in the country. Unfortunately, as the population grows, a huge number of infants and children surging every day and month to receive their prescribed vaccines. This situation may cause vaccine stock issues in barangay health centers and cities. In this paper, the researcher proposed the implementation of a DSS named Negros_Bakuna and integrate the time series method of data modeling using Autoregressive, Integrated, and Moving Average (ARIMA) to forecast the vaccine supplies and straight-line method for continual growth. As a result, Negros_Bakuna shows a 2.83 rating out of 3 during the system dry-run participated by health workers, nurses, and doctors of the pilot Local Government Unit (LGU) in the province and three experts in the domain. The ARIMA forecasting model shows 94% in predicting the next 30 observations in the time series based on the Mean Absolute Percentage Error (MAPE)of 12.67%. This paper discusses various methods and validation used in modeling the data as well as its integration on the DSS. Memores: A Web-based Intelligent Screening Tool for Predicting Social Anxiety Disorder Using A Machine Learning Model Hans Adey Cesa, Joshua Daniel Lawsin, Carlos Julian Chiongbian, Maria Angelica Zaragoza, Glenn Pepito and Rhoda Alvarez Department of Computer, Information Sciences, and Mathematics, School of Arts ans Sciences, University of San Carlos, Cebu City, Philippines 10:20-10:40 Abstract: Social Anxiety Disorder (SAD) is a common type of 1412 anxiety disorder that can disrupt a person’s work and social life. People with social anxiety disorder experience the symptoms of fear and anxiety in situations where there is a possibility of being scrutinized or judged by others. Clinicians conduct diagnostic or clinical interviews and screening tests when the patient seeks medical attention. However, the process usually takes two to four hours to complete to avoid false positive impressions leading to misdiagnosis. A study correctly detected only 2.2%, meaning the misdiagnosis rate reached 97.8%. 53
2022 International Conference on Information Technology Education (ICITE 2022) This study sought to streamline the screening process without compromising the accuracy rate of predicting a possible manifestation of social anxiety disorder within patients through an intelligent web-based screening tool that uses established machine learning algorithms to screen and evaluate a patient. The tool assesses the possibility of SAD manifestation based on the information gathered about the patient's demographic, emotional and physical symptoms and social situations. The researchers compared four machine learning models, namely: Decision Tree, Logistic Regression, Support Vector Machine, and K-Nearest Neighbors, to predict SAD. The models were trained and tested using a pre-processed dataset and then cross-validated. Using K-Fold Cross-Validation, the researchers evaluated four machine learning models based on accuracy, precision, recall, and f1 scores. The SVM model performed the best among the other models and garnered the highest accuracy of 96.01%, with 97.13% precision, 95.33% recall, and 96.13% f1 score. The researchers then integrated the SVM model into the screening tool developed. Enhancement of Offline Signature Verification Using Support Vector Machine John Ullyses Valleroso, Sherwin Carlo Cruz, Mark Christopher Blanco, Dan Michael Cortez, Khatalyn Mata and Richard Regala 10:40-11:00 Abstract: Signatures are commonly used for personal 243 verification or signifying an agreement. Therefore, confirming whether a signature is genuine has become a priority in information security. Signature verification methods are continually developed to automate and hasten the process. A study utilized Support Vector Machine to verify offline signatures. This paper proposed an enhancement of the said algorithm by adding preprocessing techniques, using a more robust set of feature extraction methods, utilizing a writer- dependent classifier, and computing the best hyperparameter values of C and Gamma using the Giza Pyramid Construction Algorithm. Two datasets of offline signatures were used to train and test the algorithms. 54
2022 International Conference on Information Technology Education (ICITE 2022) Results showed that the enhanced algorithm significantly performed better than the existing algorithm, with the 11:00-11:20 highest accuracy score of 96% and the lowest FAR and FRR 6049 both having 4%, on the same dataset with 16G of training samples. In comparison to the initial algorithm that performed the best in another dataset with an accuracy score of 83%, FAR of 17%, and FRR of 16% also with 16G of training samples. Sentiment Analysis on Student’s usage in Online Video Conference Platform using Support Vector Machines Francis Noel I. Alarcon, University of Santo Tomas – Manila Reynalen C. Justo, Laguna State Polytechnic University Sherry Mae R. Llandelar, University of Santo Tomas – Legazpi Lizel Rose Q. Natividad, San Beda University Ronel F. Ramos, FEU Institute of Technology Roda N. Sanares, De La Salle University - Dasmarinas Abstract: Since the pandemic, there has been a shift from face-to-face classes to online classes that utilize video conferencing platforms to facilitate effective communication and interaction among educators and learners. Students' perception reflects the relevant challenges they experienced due to the rapid implementation of this technology when using different video conferencing platforms for digital learning. This research paper aims to determine the student's sentiments regarding using video conferencing platforms in online classes and evaluate the learner's view of the tools used in the virtual class discussion using Support vector machines (SVMs). Survey questionnaires were distributed randomly to 200 college students currently enrolled in BSIT and BSCS programs from different colleges and universities. The researchers used a mixed-method research design, which collects numerical data that can be measured or classified using statistical analysis and employs a descriptive study design to assess the current situation and the behavior of its respondents. 55
2022 International Conference on Information Technology Education (ICITE 2022) Authors' Presentations Session VII Theme: Information Technology Time: 08:00-11:20 Venue: Room 314 (online) Session Chair: Prof. Raquel Bermudez Dr. Joey Suba Land Valuation System Using Multiple Linear Regression Crisanto Alanes 08:00-08:20 Abstract: Multiple Linear Regression (MLR) is a predictive 4794 analytical method that enables prediction using the conditional expected value of dependent variable given the values of some independent variables. This paper presents a land valuation system prototype called Land Valuation System using Multiple Linear Regression (LRLVS). The developed LRLVS intends to provide Social Housing Finance Corporation (SHFC) with an advanced tool to predict the land appraisal value using three independent variables (asking price, land area, and year) with the use of SHFC's historical data. The study also talks bout the traditional MLR method and Matrix Inversion technique applied to LRLVS model with very high accuracy in the prediction. An Energy Efficient in Routers Routing Algorithm Based on Ant Colony Optimization Jeffrey Evangelista and George Granados 08:20-08:40 Abstract: Internet is a huge connection of networks of networks 8670 around the world. It is an architecture that allows communications and methods of commerce using various computer networks around the globe to interconnect. In 2023 the global population will surpass the number of devices connected to the IP networks. There will be a 5.3 billion total internet users by 2023. The constant growing number of users and infrastructures are the main factor in the increase of power demand. One of the key players of the internet and the rest of technology are the routers. Router’s routing protocol must deal with limitation such as energy efficiency and high error rates. Using Ant Colony Optimization as the routing protocol can be lessen the power consumption of a router routing process. 56
2022 International Conference on Information Technology Education (ICITE 2022) A simulation experiments in NS2 was conducted using the following performance parameters: residual energy, packet delivery ratio and throughput. In the results shows that ACO based routing protocol performed best in the large number of routers compared with AODV and DSR. It shows also in the result that ACO based routing protocol used minimal energy in the routing process especially in the large number of nodes. ACO based routing protocol can be further extended by comparing the other bio inspired algorithm where learning approaches have been performed like ants. Alab: An E-learning 2D Platform Game on Readings in Philippine History Using a Priority Queue Algorithm Kit Daniel Jay T. Lim and Christine F. Pena University of San Carlos, School of Arts and Sciences, Department of Computer, Information Sciences and Mathematics, Cebu City, Philippines 08:40-9:00 Abstract: Philippine History is an important field of learning for 647 Filipino citizens and communities. With the hopes of promoting such, an e-learning 2D platform game was developed that can 9:00-9:20 serve as a supplementary learning tool towards the General 424 Education course, Readings in Philippine History (GE-RPH). The 2D game was composed of four stages depicting specific events during the Pre-Colonial era. An algorithm that uses a Priority Queue to determine how in-game questions are selected from the pool of questions was developed. From the population of 35 students, it is shown that about 85.8% of students found the e- learning application useful for expounding their knowledge in GE-RPH. ALERTO: An Android-based Accident Hotspot Tracker Notification System Utilizing Analytics and GPS Tracking Algorithm Inocencio C. Madriaga Jr., Caryl Anne D. Elisterio, Mark Joseph S. Caisip and Junnel A. Ramirez Batangas State University ARASOF Nasugbu Abstract: Road traffic accidents are now considered as the top leading cause of death worldwide that is unrelated to any medical diseases. 57
2022 International Conference on Information Technology Education (ICITE 2022) Furthermore, CALABARZON, had the most traffic injuries in 2018, accounting for 16.5 percent of all car accidents 9:20-9:40 nationwide. Thus, the objective of this study is to develop an 863 application that will notify the user when they are 200 meters near or heading to an accident hotspot, this will help them to be more cautious and attentive when traversing different routes of Nasugbu. ALERTO is a mobile application that provides information about the different accident hotspots in the municipality and alerts the user about them. RAD and K-means clustering were applied to the development of the study. The researchers used the RAD methodology since it is efficient to use for a shorter time frame for it focuses on prototyping and feedback rather than planning, so the development is faster. K- means clustering was used in grouping the accident hotspots in the different areas of Nasugbu. A survey was used to evaluate the system. After the evaluation, the gathered data were tallied, computed, and interpreted. Based on the results, most of the accidents in Nasugbu happen along the National Road of Barangay Lumbangan, Bilaran, and Banilad. The absence of road signages and without thoroughly checking the vehicle condition are the leading causes of the occurrence of an accident in the municipality. Respondents highly accept the mobile and web applications and strongly agree with their effectiveness. Innovated Smart Attendance System with Face Recognition and Temperature Sensing Features Christopher E. Barientos, Chester L. Cofino, Jun -Jun A. Obiso Abstract: The research aims to develop an innovative smart attendance system to address issues from using the current biometric attendance system. It focuses on developing an improved attendance system with face recognition technology with thermal sensors that will capture individual data in a contactless manner. The researcher used structured planning and designing an approach to develop a system with a face access device equipped with Artificial Intelligence (AI) face recognition capturing data up to three meters maximum distance. It comprises a face access device, the HR Web App, and a mobile or web application accessible anytime and anywhere within reach of a wi-fi connection inside the campus. 58
2022 International Conference on Information Technology Education (ICITE 2022) The study is anchored on the Automata Theory, Control Theory, Theory of Innovation, and two legal basses under CSC Exec. Order 292 and Accessibility Law of 1982. After its development, the prototype was installed and tested at the gate. A survey questionnaire was used to determine the effectiveness and assessment level of the system as rated by the users and experts after testing the prototype. The study found that both experts and users recommend the system for implementation as they believe that the system is very effective in monitoring the attendance of employees, convenient, and facilitates ease for the Human Resource department in processing, generating, and printing of log information. Furthermore, this innovation can still be improved, such as upscaling the battery capacity, availability of a dedicated data server account, and a university policy for its secured utilization. QR Authentication Security System Prof. Rizalina C. Valencia, Nathan Jethro C. Santiago and Sean Alexis D. Escallar 9:40-10:00 Abstract: The purpose of this thesis/system is to innovate the 6800 login process for any kind of system that holds asset or any kind of important data’s. By combining different kind of authentication such us the username and password authentication, an OTP (One Time Password), and encryption of QR code we can add another wall of security for the company and the beneficiary. We also add a web application wherein customers and consumers can view the services that the company can provide and view ways of contacting the company. Another purpose of the web application is to provide a unique feature for the administrators of the company. The Administrator can Login in the web application where it directs them to the dashboard. The dashboard provides the Creation, Modification and Deletion of accounts. It also can show the existing account in the database and also shows the time in and time out of the employees. In addition the dashboard can also show the concerns of the employees to their account. Since we are using 2 different type of application (Desktop Application and Web Application) we need to use a single type of database that can support the two. The best choice is to use a cloud database since we have to create a web application. 59
2022 International Conference on Information Technology Education (ICITE 2022) To the future researchers the system may have lots of flaws in the design and some additional features are missing but the main objective of our research is met, it is functioning well and have a lots of room for improvement. We hope that this research can contribute in the security of any kind of application and non-stop innovation of technology. E-skedyul: An E-government Financial and Goods Assistance Transparency Website Francesca Marie Alcaraz, Ace Christian Aruego, Czesca Louise Dizon, Zhuntelle Nicole Manalo, Leonard Alejandro, Marvi Aresta-Bayrante and Ma. Carmela Racelis Information Technology and Information Systems Department, Adamson University, Ermita, Manila, Philippines 10:00-10:20 Abstract: The goal of this study entitled E-Skedyul: An E- 9576 Government Financial and Goods Transparency Assistance Website is to close the gaps in financial aid and commodities distribution in the community with speed, transparency, and efficiency. The study involves developing three primary aspects of the system; monitoring, tracking, and request that will make claiming of financial aid or “ayuda” easier and hassle free. This study utilized quantitative research methodology where respondents from the Office of Senior Citizens Affairs (OSCA) of Manila City were chosen using purposive sampling. The data gathering procedures included interview, observations, and surveys. Agile Methodology was used in the development of the system. The beneficiary company was visited frequently for constant collaboration with stakeholders and continuous improvement at every stage. The ISO 25010 Quality Model for Web Applications was utilized in evaluating the software. The system was presented to the target users so they could assess it for functional suitability, performance efficiency, compatibility, usability, reliability, security, maintainability, and probability. The findings of the initial testing showed that the feature of proposed system suits its intended purpose and garnered a rating of 4.61 interpreted as \"Very Good\". 60
2022 International Conference on Information Technology Education (ICITE 2022) 10:20-10:40 Hiraya: Pivoting the Philippine Animation Through 3076 Integration of 2D Cut-out Animation and 3D Asset Roselle S. Basa, Faculty, College of Computer Studies and 10:40-11:00 Systems, University of the East Manila 7214 Ariana Gabrielle N. Bentir, Brdiget Anne R. Espiritu and Earl Dominic G. Guevara, Alumni, College of Computer Studies and Systems, University of the East Manila Abstract: This 2018 animated short, entitled, Hiraya, showcased a thriller story while highlighting various Filipino cultural references using the latest trend in animation. It tells a tale of a caring nanny, Ate Merly, and her ward Bobby. Despite being born with glaucoma, Bobby is a talented artist, who spends most of his days painting – often about his dreams. With parents who are frequently unavailable because of work, Bobby is usually accompanied only by his bubbly nanny named, Ate Merly. Their monotonous daily life suddenly takes a turn when unexplained and paranormal events suddenly started happening with Bobby. The animators were inspired by Tim Burton’s approach in designing the characters of the film. Using Toon Boom Harmony, the characters came to life through cut-out animation. These were integrated with three-dimensional environments modeled and rendered using Autodesk Maya. The eerie-feel was achieved by using techniques in horror and thriller films such as Dutch camera angle, handheld camera shot, and sound effects with extreme frequencies. Softwares namely Clip Paint Studio, Adobe Premiere Pro, Adobe Photoshop, and Adobe Audition were also utilized to bring thestory of Hiraya to life. SCVAX: A Solution to Nationwide Vaccination Management Information System Ritchie Reyna, Darwin Mangca, Alma Christie Reyna, Jovie Gallera, Perfecto Ruaya, Ghandi Galila, Ralph Aran Cabanero, Antonio Jr. Piloton and Arvin Salvador Abstract: SCVax is a powerful application platform that can be used to locate and monitor the total population of vaccinated and unvaccinated individuals in a certain community. 61
2022 International Conference on Information Technology Education (ICITE 2022) This paper summarizes various and unique features in the field of mapping and visualization in terms of handling and 11:00-11:20 interpreting data collected from the system, it also 2053 provides reliable information, and secures the confidentiality of the data with the integration of the QR code function. The main focus is to design, develop and deploy a comprehensive and efficient system application to provide a real-time vaccination validation and certificate to lessen the use of fake vaccination cards and to provide travel authorities with vaccination passport features following DOH and IATF guidelines. Data gathering is the vital method used in the initialization of the system and the Codelgniter and PHP framework with built-in protection against CSRF and XSS attacks and is compatible with a Model-View-Controller (MVC) design pattern and industry-standard web development of scalable projects used in the design and development of SCVax. The platform is capable of generating government- mandated reports requirements concerning the COVID-19 vaccination program. The use of SCVax is more accurate and faster compared to the manual registration processes and it can support ongoing contract tracing as a vaccination QE code is used and also generate health- related statistics reports during a disaster which can identify if an individual is vaccinated or unvaccinated upon entering an establishment. A researcher-made questionnaire based on applicable software characteristics was used in system evaluation. It was done by taking the responses from respondents who used the system. The system was rated favorably by users in functionality (4.57), usability (4.48), security (4.43), and portability (4.44). With an average rating of 4.6/5, SCVax was found to satisfy the requirements. A Data Mining Approach Towards the Absorption of Student Interns to Host Training Establishments Arnold B. Galve and Mengvi P. Gatpandan Abstract: The primary goal of internship is for students to gain as much experience as possible in the environment of Host Training Establishments. 62
2022 International Conference on Information Technology Education (ICITE 2022) CHED mandates Higher Education Institutions to provide students with internship opportunities to apply their knowledge and skills to practical workarounds of the workplace. The collaboration between HTEs and HEIs to provide student interns impactful work experience is essential to internship success. The study involved selected private HEIs in wherein a total of 1,019 instances were used as dataset with 9 predictors extracted from OJT evaluation sheets obtained from the said HEIs. An ensemble of binary classification data mining algorithms was used to discover patterns of internship absorption to HTEs. Results of the binary classification models revealed accuracy scores of 100% for decision tree, 83.78% for KNN, 81.08 for logistic regression and support vector machine and 64.86% for naïve bayes which means that the independent variables used are significant predictors of internship absorption to HTEs most notably the student personality, technical knowledge & skills, office work management and GPA. HEIs can benchmark on these results and include the predictors used in this research to craft or revise their internship evaluation criteria to improve the oversight of their internship program. 63
Search
