Analytics@TP Pre resen ented ed by: : Michael Yap 2018-09-28
Agenda Our Analytics Journey • Capability Development • Challenges • Sample of Data Products • Student Analytics • Learning Analytics • Graduate Analytics • Procurement Analytics • Text Analytics • IoT Analytics • Summary • 2018-09-18 2
Our Analyt ytics Jo Journey 2018-09-18 3
Gartner Analytics Ascendancy Model Raw Data & Reports Data Information Knowledge Insight Action 4
What can we analyse ? Manpower Optimisation Learning Engagement Talent Mgmt CCA Internship / OCP HR Attrition Attendance Capability Development Attrition Alumni Staff Acad Performance Students Financial Assistance/ Awards Graduation Admission Procurement Finance Estates & Facilities Industry Relation IT Customer Satisfaction 5
TP Analytics Roadmap Social media Alumni Outreach Analytics Finance Analytics Analytics Analytics Utility Data Scientist Analytics Industry Training Student Partner Support Analytics Analytics IT Resource Procurement HR Analytics Analytics Learning Data Analyst Analytics Analytics Training Graduate Awareness Student Analytics Visual Analytics Training Analytics Training 6
Early Learning Analytics Project • Developed and launched in 2014 • A front-end self-service analytics tool for School Directors and Course Manager to gain academic insights – on course and subject performance 7
Capability Development 2018-09-18 8
Capability Development • St Start fr from basi asic : : Awareness • Clas lassroom le lear arnin ing • eLe Learning • On On-the-job-train inin ing • Certific ication • Community of of Practice (C (CoP) • Knowledge Sh Sharin ing 9
Challenges 2018-09-18 10
Challenges Data sources and Quality • Str Structured vs s unstructured data • Da Data consis istency • Da Data qual alit ity is is im important in in producin ing meanin ingful l resu sults Competency Building • Steep Le Learnin ing Curve • La Lack of of sk skill illed personnel l in in busi siness an anal alyt ytic ics • Coll ollaboration wit ith dom omain in experts an and IT IT ap appli lication teams Familiarisation with tools • Ne New an anal alyt ytic ical l tools an and systems • Di Different tools ls for dif ifferent role les - Bac ackend, fr frontend, Admin inistration 11
Challenges Change Management • Di Different min indset – data-driven decis ision makin ing • Str Strategic vs s op operation • De Descriptive to o Predictive an analyt lytics System Performance • Reasonable loa loadin ing an and resp sponse tim time • Dr Drill ill down, drill rill th through • Da Data si size doe oes matter Access Control • Di Different fr from tr transaction system • Aggregated data • Open for se self lf-servic ice 12
Sample of Data Products Student Analyt lytics 2018-09-18 13
Conceptual Architecture Data Sources Data Visualisation Business Analytics Management Users • Pre-Poly Data Learning Analytics • Academic Extract, Performance Descriptive Transform Student Analytics • Student Analytics Load Demographics • Attendance Insight Procurement Analytics • Learning Periodic Management MS SQL refresh System • CCA • GeBIZ • Finance Data Predictive Graduate Analytics Analytics • Graduate Employment Top Performer / Data Marts Foresight At-Risk students Oracle, MS SQL Files SQL Server Integration Services SAS Visual Analytics SAS Data Integrator SAS Enterprise Miner 14
Student Analytics • Student Academic Performance • Reports for BOE (Board of Examiners) • Graduation & Attrition • Comparison by Admission Category / Entry Qualification • 5-year Trend • Comparison by Predictive Analytics – Top Performer / At Risks 15
School BOE Report Sch C Sch A Sch D Sch E Sch F Sch B AAA BBB CCC DDD EEE FFF GGG HHH 16
Graduation/Attrition Report 17
Admission Category 18
Entry Qualification 19
Predictive Analytics Predictive models were built 20
Sample of Data Products Learning Analyt ytics 2018-09-18 21
Learning Analytics • Support Learning Intervention and Enculturate Reflective Practice • How long did students engage with online content? • What is the level of student engagement in the online discussion forum? 22
LMS Content Access Filters Students’ Access Patterns 23
Student Workload Distribution 24
Sample of Data Products Graduate Analyt ytics 2018-09-18 25
Graduate Analytics To distil key drivers for graduates’ outlook to enable personalised interventions Predict Distil underlying Enable Intervention Graduates’ Outlook (every semester or as needed) Key Drivers Will the graduate be Generate propensity of students economically active? Demographic at the end of each semester for working in field related to studies? Entry Qualification ‘personalised’ interventions engaged in further study? Academic Performance etc. Financial etc. 26
Input Considerations Utilise data in TP source systems to study students’ behaviour comprehensively Entry Academic Disciplinary / Demographics Financial Non-academic Qualification Performance Attendance • Age • Admission • Core / Elective / • Award / Bursary • CCA points • Disciplinary Category Group CDS Subjects Record • Citizenship • PCI Range passed / failed • Choice Order • Exam MC • Gender • etc. • GPA • Entry • Leave Days • etc. Qualification • Subject Marks • etc. • etc. • etc. 27
Model Building Score Model* Data Data Data Code (Decision Loading Partition Scoring Tree) Export • Deployment • • • • Historical data Configure the % Determine Run the model of code • Drop of training and model(s) with the with the unnecessary validation data best balanced validation data variables outcome 28
Key Modelling Methodology Decision Tree widely used by organisations for its intuitiveness and business interpretability For Illustration: All Students Job Market Leakage Model (extract) Job Market Leakage: 32% High-risk groups Others: 68% Lower-risk groups School A, … School School B, … Job Market Leakage: 26% Job Market Leakage : 40% Others: 74% Others: 60% 17.5 X Gender Y < 17.5 O-Level Raw Aggregate Job Market Leakage : 36% Job Market Leakage : 21% Job Market Leakage : 32% Job Market Leakage : 49% Others: 64% Others: 79% Others: 68% Others: 51% <3 GPA >=3 Job Market Leakage : 35% Job Market Leakage : 21% [For illustration only] Others: 65% Others: 79% 29
Sample of Data Products Procurement Analyt ytics 2018-09-18 30
Procurement Analytics • • No response or single response ? Specification Frequency of purchase for specific item ? too stringent or geared towards a particular Spending patterns ? brand of item? 32
Procurement Analytics – Alerts & Audit • Alert functionality - prompt relevant stakeholders to review the data so that necessary intervention can be considered at different stages of the procurement process. • Apart from intervention, information gathered can also be used for audit function. 33
Text xt Analyt ytics 2018-09-18 34
Text Analytics - Project Background • TP conducts various surveys – Teaching Effectiveness, Subject Review, Course Review • Extensive analysis of quantitative data • Eyeball qualitative data – E.g. Online Student Evaluation of Teaching (OnSET) collects about 100,000 free-text comments annually Objective Leverage on technology to analyse free-text comments, so as to gain insights on themes and sentiments
Key Advantage: Categorization 36
Benefits of Text Analytics • Convert open-ended comments into meaningful themes and quantifiable results • Automate comment processing, saving time and resources • Leverage on the purpose- built ‘Teaching and Learning Dictionary’ • Obtain a more complete picture of what students are saying 37
Text Analytics • Cross-tabulation: Qualitative & Quantitative • Provide better insights Q11: Write down something that your lecturer has done especially well Female students reflect more positively on the learning experience. Key insights for further research: Are there more female faculty? Are there more female students? What is the graduation rate for females? What is the employability rate for females? 38
Io IoT Analyt ytics 2018-09-18 39
Video Analytics Car Plate Number Recognition 40
Video Analytics Using CCTVs and video analytics for people counting. 41
Analytics with IoT Sensors • Environment – Temperature, CO2, PMI sensors • Carpark Occupation/Utilisation – carpark sensors • Energy Management – Smart Distribution Box • etc 2018-09-18 42
Summary ry 2018-09-18 43
Summary Our journey continues… 2018-09-18 44
Thank you ! 2018-09-18 45
Recommend
More recommend
