DEPARTMENT OF COMPUTER SCIENCE SCIENCES CLUSTER UNIVERSITY OF THE PHILIPPINES CEBU
DCS Research Direc-ons Prepared by: Asst. Prof. Kurt Junshean Espinosa Chair Harolds Hotel, Cebu City March 19, 2014
DCS Vision A world-‐class insRtuRon in computer science research and educaRon that shapes and transforms the region into a dynamic and ICT-‐driven society Mission 1. To advance knowledge in computer science through novel and innovaRve interdisciplinary research that improves the quality of life and society 2. To produce graduates who are technically competent and socially responsible leaders in CS research and in the ICT industry 3. To opRmize the use of ICT by strengthening partnership with the different stakeholders
History • 1996 – started as a BS Computer Science Program • Mar 2007 – insRtuted as Department of Computer Science • Dec 2007 – awarded by CHED as Center of Excellence in IT EducaRon (1 of the 9 in the Philippines) • 2009 – started the MS Computer Science program
DCS • People: 1 PhD, 5 MS, 2 MS Cand, 2 instructors, 4-‐6 lecturers (from: IBM, Lexmark, NCR, NEC, etc), 1 visiRng professor (Austria), 3 affiliate professors (UP Diliman) • Curriculum: Based on ACM/IEEE RecommendaRons and stakeholders’ input • Career Path of Graduates: academicians and researchers in universiRes and R&D companies, sofware engineers, startup entrepreneurs, IT managers • Research Areas: Machine Learning and Big Data AnalyRcs, Natural Language Processing, Digital Signal Processing (audio, video, image), E-‐learning, OpRmizaRon and HeurisRcs, Social Network Analysis, Computer SimulaRon and Modeling
Funded researches
1. ASEAN Machine TranslaRon (MT) • The project aims to provide a public service of machine translaRon for all ASEAN languages. • Supports opening of ASEAN Economic IntegraRon by 2015 • aseanmt.org • Faculty: Prof. Robert Roxas • SpecializaRon: Natural Language Processing (NLP) • Fund Source: CHED • Collaborator: DLSU Manila, Center for Language Technologies
2. DREAM Program • Disaster Risk and Exposure Assessment for MiRgaRon (DREAM) Program • Covers areas most hit by floods • Faculty: For UP Cebu, Dr. Sinogaya and Prof. Silapan together with DCS faculty (Hazard Mapping, Assessment of AquaRc Resources) • Fund Source: DOST • Collaborators: UP Diliman Eng, NIGS
2. DREAM Program: LIDAR • Light DetecRon and Ranging (LiDaR) • Captures topographic data
3. Dengue Epidemic Modeling • The first aim is to support decision makers by esRmaRng the outcome of intervenRons against dengue like: – mosquito control – increased usage of repellents or – vaccinaRon strategies [vaccines are currently under development and are expected to be released soon, maybe already 2015]
Source: hmp://upload.wikimedia.org/ wikipedia/commons/4/49/Dengue06.png
3. Dengue Epidemic Modeling • The second aim is to provide more insights into the spreading process of dengue among a populaRon. • The vision is to help understanding dynamics of dengue epidemics bemer, and even to be able to make forecasts on outbreaks one day.
3. Dengue Epidemic Modeling • Faculty: Dr. Florian Miksch, Prof. Kurt Espinosa • Fund Source: UP Cebu Research Grant • SpecializaRons: Computer SimulaRon and Modeling, Machine Learning • Collaborators: DOH Region 7 and Cebu City Health Office (possibly: George Mason University under USAID STRIDE fund) • Pilot project is being done now for Cebu City and based on the results possibly extend the model for enRre Philippines in 2015.
4.Crime Modeling and PredicRon • Title: “Mapping and Modeling of Crime Incidences in Cebu City to be used in Crime PredicRon for Public Safety” • Collaborator: Cebu City Police Office
4.Crime Modeling and PredicRon • Faculty: Mr. Ryan Dulaca (Project Lead), Prof. Kurt Junshean Espinosa, Prof. Demelo Lao, Mr. Ronald Pernia • Areas of specializaRon: Machine Learning, Social Theory, StaRsRcs • Fund Source: Center for IntegraRve and Development Studies (CIDS)
5. Automated Image ClassificaRon • Title: “An automated viral infecRon classifier in a Carabao mango tree” • Collaborator: Department of Agriculture • Faculty: Prof. Sandra Famador • Fund Source: UP Cebu Research Grant • Area of SpecializaRon: Digital Image Processing
hmp://www.yspuniversity.ac.in/eap/ images/res-‐26.jpg
NON-‐FUNDED RESEARCHES
Prof. Kurt Junshean Espinosa
SOCIAL NETWORK MINING
1. Intelligent Tutoring System • How can Social Networks be tapped to learn a new language?
1. Intelligent Tutoring System • Faculty: Prof. Kurt Espinosa, Prof. Jaime Caro • Research Area: Intelligent Tutoring System, Machine Learning • Collaborators: Department of InformaRcs, University of Piraeus, Greece (Prof. Maria Virvou and Christos Troussas) • Has published journals and conference proceedings.
2. Opinion Mining • How much data in social network can reveal people’s senRments on parRcular products (or enRRes in general)? • Is it possible to summarize senRments on parRcular product components?
3. Disaster-‐related topics • Can we infer the actual locaRon of a parRcular disaster by studying the community talking about it? – “Community Structure DetecRon and Analysis in Disaster-‐Related Tweets”*
• Can we detect where disasters are happening thru the social network feeds? – “ParRcipant Tweet IdenRficaRon using Support Vector Machines”* • Accepted for publicaRon in Philippine CompuRng Journal
4. EmoRon Diffusion in Social Networks • How does an emoRve post/status spread throughout the network? • What factors characterize emoRon diffusion in social networks?
E-‐LEARNING
1. InteracRve Learning Object • Protein Synthesis ILO (thru CILOB) • Collaborator: Ms. Jeraline Gumalalm, UP High School Biology teacher • Faculty: Prof. Demelo Lao
2. InteracRve SimulaRon Engine • “How can we simulate events in a more dynamic and interacRve way?” • Collaborator: UP Diliman DCS (Prof. Jaime Caro) • Faculty: Prof. Espinosa
Mr. Jedaiah Joel Lumagbas
CS EDUCATION
1. Technology and EducaRon l
l
How can technology be used to enhance learning experience? Faculty: Mr. Jedaiah Lumagbas (currently doing MS at University of Melbourne)
2. Serious Games • How can games be designed to improved learning exp?
Prof. Paula Esplanada
OPTIMIZATION AND HEURISTICS
1. OpRmizaRon Algorithms l l l
l
GeneRc algorithm Simulated annealing Ant colony opRmizaRon Example Problem Area: How to schedule classes for opRmal use of resources?
hmps://encrypted-‐tbn1.gstaRc.com/images? q=tbn:ANd9GcQZPtUNHIpLAWtQBqIqe2stbfXyoGPTbiAxxaPvp J56yT-‐BxG4g8Q
Any quesRons?
THANK YOU FOR LISTENING