MCITE 2018 Abstracts

1984 + 34: Big Data *Finally* Killed the WWW Star

"If the Internet stumbles, it will not be because we lack for technology, vision, or motivation. 
It will be because we cannot set a direction and march collectively into the future."

Coordinated Aerial Communication System for Emergency Situations through Wireless Communication Devices

Raul Vincent W. Lumapas, MSc
Ateneo de Davao University

Abstract.  The emerging use of Unmanned Aerial Vehicle (UAV) in different applications, both civic and commercial, opens the possibility of its usage in far more critical applications such as disasters and emergency situations. This study explored the possibility of an air-to-ground (a2g) / ground-to-air (g2a) ad hoc communication paired up with a mesh network implementation of mobile devices utilizing its network capabilities, such as the IEEE 802.11n standard. The study utilized Unmanned Aerial Vehicles for the Aerial Ad Hoc network and uses a standard implementation of a wireless mesh network for the mobile devices. This study presented different schemes and was tested and simulated to identify the efficiency of the schemes developed for this study. Furthermore, it identified the scheme’s applicability in emergency situations communication. After series of tests, it was determined that a2g/g2a communication is viable and can be further be improved. The communication capabilities of the UAV may be limited, such as in its capacity, but can be further be improved or a separate external communication point, such as an external access point, can be utilized. The possible schemes developed is proven to be helpful in proper utilization of the communication capabilities provided by the UAV or the external network component attached to the UAV. The test had provided the distance ranges wherein proper communication can be established and utilized, be it line-of-sight or out-of-sight due to obstructions as a result of disasters.

The effects of Scaffolding to various Prior Knowledge Learners in a Learning by Teaching Environment


Cristina E. DUMDUMAYA1,2, Ma. Mercedes T. RODRIGO1
1Ateneo de Manila University, Philippines
2Institute of Computing, University of Southeastern Philippines, Philippines


Abstract.  We compared the effects of cognitive and metacognitive scaffolding and prior knowledge on students’ performance within a learning-by-teaching intelligent tutor for algebra to investigate factors that mediate students’ learning. Findings revealed that both cognitive and metacognitive scaffolding increased learners’ achievement however, no significant difference between conditions was detected. Analysis on the learning gains of high prior ability and low prior ability students, given cognitive scaffolding, revealed no significant difference, while metacognitive scaffolding facilitated learning for low prior ability learners but obstructed the performance of high prior ability students. Further investigation on what might cause low prior ability learners to gain more from metacognitive scaffolding than high prior knowledge learners revealed that the hint avoidance behaviour exhibited by students negatively correlates to the learning gains of high ability learners. Moreover, metacognitive actions related to self-regulatory learning’s reflection sub-function correlates negatively to the learning outcomes of low ability students.

Predicting Student Carefulness within an Educational Game for Physics using Support Vector Machines

Michelle P. BANAWANa,b, Ma. Mercedes T. RODRIGObJuan Miguel L. ANDRESc,
aAteneo de Davao University, Davao City, Philippines
bAteneo de Manila University, Quezon City, Philippines
cUniversity of Pennsylvania, Philadelphia, USA

This paper can be found in Chen, W. et al. (Eds.) (2017). Proceedings of the 25th International Conference on Computers in Education. New Zealand: Asia-Pacific Society for Computers in Education.  Click to access the full paper

Abstract. Student carefulness is defined as being attentive, mindful or focused on the task at hand. In this paper, we create a predictive model for student carefulness within an educational game called Physics Playground (PP). We used game logs and manually-labeled gameplay clips of 54 students from the Philippines to develop three support vector regression models that predict carefulness using: (1) predictors of the game developers, (2) predictors from social science research, and (3) the combination of these predictors. After preprocessing and feature selection, the support vector regression models were able to significantly predict student carefulness. This research’ empirical findings suggest that carefulness in Physics Playground can best be predicted by expanding the model of the game developers and including predictors that have been previously researched in the broader social science literature.

Keywords: carefulness · machine learning · support vector machines · regression· Physics Playground

Detecting Ransomware in Android using Machine Learning 
Oneil B. Victoriano, Ateneo de Davao University, Doctor in IT (DIT) Student

Abstract: Android market share increase significantly from less than 10% in 2009 to 85% in 2016. With Android domination in the mobile Operating System wars, most of the malwares are also detected in the said OS. Ransomware is a type of malware wherein the malicious code locks the device, encrypts files, and demands payment from the victim to unlock the device and data. The Ransomware detection reports from cybersecurity companies triggers high threat in Android devices vulnerability. The study used machine learning approaches, particularly classifiers: Decision Tree, Random Forest, Gradient Boosting Decision Trees, and AdaBoost to detect Ransomware malware. The study used dataset from HelDroid study with known Ransomwares features, the dataset was transformed and feed on the classifier model. Using 5- attribute dataset feed on the classifier, the models generate high average of 98.05% accuracy rate, both on training and test sets. Feeding the binarized 229-attribute dataset, Decision Tree generates 99.08% accuracy

LeafCheckIT: A Banana Leaf Analyzer for Identifying

Macronutrient Deficiency


Jonilyn A. Tejada* and Glenn Paul P. Gara**

*Davao del Norte State College, New Visayas, Panabo City

**University of the Immaculate Conception, Davao City

This paper is published at the ICCIP '17 Proceedings of the 3rd International 
Conference on Communication and Information Processing. Pages 458-463.

Abstract.  Nutrient deficiency affects the production of banana fruits. Plant malnutrition may visually reflect on their leaves, however, identifying deficiency symptoms could be a difficult task and often requires laboratory tests such as leaf and soil analysis. This study introduces the LeafCheckIT, a web and mobile application that uses Random Forest machine learning algorithm to identify Nitrogen (N), Phosphorus (P), and Potassium (K) collectively known as macronutrients deficiency symptoms on banana leaves. Based on the training set evaluation and 10-fold cross-validation test conducted on WEKA data mining software, the technique used in the application resulted in 100 and 91.64 percent accuracy rate respectively.

Application of Support Vector Machine (SVM) and Quick Unbiased Efficient Statistical Tree (QUEST) 

Algorithms on Mangrove and Agricultural Resource Mapping using LiDAR Data Sets

Carmel Jean G. Madanguita, John Paul L. Oñez, Herzon G. Tan, Milce D. Villanueva, Jesson E. Ordaneza, Remie M. Aurelio Jrb., Annabelle U. Noveroc
Phil-LiDAR 2 Project, College of Science and Mathematics, University of the Philippines Mindanao, Mintal, Tugbok District, Davao City, Philippines

This paper is one of the outputs of the LIDAR2, a project funded by the Department of Science and Technology
Presented at the International Congress on Chemical, Biological, and Environmental Sciences 
Published in the International Journal of Applied Environmental Sciences (IJAES)

Abstract.  Accurate mapping of mangrove and agricultural areas is necessary for effective planning and management of ecosystems and resources. While expert interpretation has been the typical method of classifying data sets, more efficient, objective, and faster methods of classification are required. This study applied the two classification techniques namely Support Vector Machine (SVM) and Quick Unbiased Efficient Statistical Tree (QUEST) algorithms for mapping mangrove and agricultural resources using LiDAR data. Ten LiDAR data sets were used for mangrove delineation. Each data set had a total of 90 ground-truth samples (30 per class) and 150 training points (50 per class) grouped into three classes: Mangroves, Other Vegetation and Non-Vegetation. Using Lastools software CHM, DSM, DTM, Intensity, Hillshade, Numret and Slope derivatives of the three LiDAR blocks were generated. eCognition software was used to perform classification of mangroves. A paired t-test was done to compare the accuracy of these two algorithms to determine which performed better in classifying mangroves. For agricultural resource mapping, LiDAR data sets for Tagum City and Panabo City were analysed. These areas contain large banana, coconut, and mango plantations.  Statistical analyses showed that SVM performed better than QUEST in mangrove delineation. In agricultural resources mapping on the other hand, results showed that SVM and QUEST combined improved the general overall accuracy for Tagum and Panabo Cities to 97% and 96%, respectively. The agricultural land cover extracted could be used for a more accurate and effective resource management and monitoring of the cities’ agricultural land. Both SVM and QUEST have a potential to improve the overall accuracy of LiDAR blocks in both mangrove and agricultural areas.

Keywords: SVM, QUEST, LiDAR, Mangrove mapping, Agricultural Resource mapping

Alternative Mass Transport System For Davao City: A Geographic Information System Approach

Mark Van M. Buladaco

Davao del Norte State College, New Visayas, Panabo City

Abstract.  Mass transportation refers to the machines which are specifically created to carry a certain number of people from one area to another. In highly progressive and populated places such as in Metro Manila Philippines, it becomes conventional among people who commute every day on their way to work, school, home and etc., for instance, in the Philippines, there are Metro Rail Transit (MRT) and Light Rail Transit (LRT) found in Metro Manila. As the number of mobile vehicles increases, problems slowly appeared that soon affected the citizens especially the commuters. Thus, it also affects the environment. The situation is also gradually appearing in Davao City, for the city itself has been having a positive progress economically. And to ease the negative effect of a sudden increase in a quantity of the mass transportation in the City, the researchers have conducted a study that was mainly focused on making an alternative mass transport system for Davao City using the FOSS Geo-Spatial Technique. This study was conducted in the city of Davao, in order to make another route with regards to the existing Korean LRT route to ease the traffic congestion on the proposed stations of the proposed Korean LRT.

The study found out that the Population of Davao City is gradually increasing per year and the implementation of Light Rail Transit is necessary to provide solution to the uprising dilemma. The commercial areas in Davao City are located in areas where the population is high and the people's flow is steady and the road network is accessible. The Road Network in Davao City is accessible to have an alternative route for Mass Transit. The Traffic Congestion in Davao City are located in the Primary Roads, Secondary roads, and National Highways and also pinpointed in commercial zones with high-density residential zones. The alternative Mass Transport System in Davao City is the combination of 2024 projected population, Traffic Congestions and the Commercial Areas in Davao City

Keywords:  Geographic Model, Mass Transport, Davao City, GIS

Characterizing Collaboration Based on Prior Knowledge in a Pair Program Tracing and Debugging Eye-Tracking Experiment

Maureen M. Villamor* and Ma. Mercedes T. Rodrigo**
*University of Southeastern Philippines, Davao City, **Ateneo de Manila University, Quezon City, Philippines

(This paper has been presented to the 10th International Conference on Educational Data Mining in China.  Click here to access the full paper)

Abstract.  We characterized the extent of collaboration of pairs of novice programmers based on prior knowledge as they traced and debugged fragments of code using cross-recurrence quantification analysis (CRQA). We built cross-recurrent plots using the eye tracking data and computed for the CRQA metrics, such as recurrence rate (RR), determinism (DET), average diagonal length (L), longest diagonal length (LMAX), entropy (ENTR), and laminarity (LAM) using the CRP toolbox for MATLAB. Findings revealed that both low prior knowledge pairs (BL) had higher RR than both high prior knowledge (BH) and mixed (M) pairs, implying that the BL pairs either collaborated better or struggled more in program comprehension and debugging. The BH pairs had shared the least identical scanpaths (lowest DET) but this could be due to more divergent episodes, implying more independent thinking. The M pairs possibly exhibited more attunement (highest L), suggesting the presence of a leader-follower tandem. The BL pairs were more stable in terms of eye coordination (highest LMAX), exhibited more complex scanpaths (highest ENTR) implying the use of trial and error strategies for locating bugs, and spent more time in certain regions of the code (highest LAM) indicating that they struggled more in program comprehension and debugging.