learning ▪ How is traditional programming different from machine learning? 2. How machine learning is changing e-government? ▪ E-government in Indonesia ▪ Automatic text classification ▪ Automatic number plate recognition ▪ Handwritten text recognition ▪ Potential benefits of machine learning 3. Challenges of Using Machine Learning in Government Text Preprocessing Pipeline for Bahasa using Python: Concept, Steps, Tools, and Examples Kuncahyo Setyo Nugroho | Present in PyCon ID 2020
Difference? Automating E-Government Services with Artificial Intelligence: Extensive use of Machine Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021
Services with Artificial Intelligence: Extensive use of Machine Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021 Use case: How can a computer recognize and separate apples and oranges?
Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021 Traditional Programming if pixel[5][8] is red & pixel[2][8] is red: if pixel[9][4][3] is green & …: return ‘apple’ … … … else: return ‘orange’ input output Rules Answer Data
Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021 Traditional Programming: How About This?
Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021 Machine Learning input output Apple Apple Apple Oranges Answer Data Rules
Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021 Machine Learning ? Apple Machine Learning Rules “Field of study that gives computers the ability to learn without being explicitly programmed” (Arthur Samuel, 1959)
Artificial Intelligence: Extensive use of Machine Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021
with citizens, businesses, and other arms of government. (United Nations, 2006) Automating E-Government Services with Artificial Intelligence: Extensive use of Machine Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021 E-Government
Tahun 2003 tentang Kebijakan dan Strategi Nasional Pengembangan E-Government. 2. Peraturan Menteri Pendayagunaan Aparatur Negara dan Reformasi Birokrasi Nomor 5 Tahun 2018 tentang Pedoman Evaluasi Sistem Pemerintahan Berbasis Elektronik. Automating E-Government Services with Artificial Intelligence: Extensive use of Machine Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021
use of Machine Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021 Hardaya, I. S., Dhini, A., & Surjandari, I. (2017, October) https:/ /monkeylearn.com/blog/customer-complaint-classification
Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021 Manwatkar, P. M., & Singh, K. R. (2015, January) Handwritten Text Recognition
leads to the generation of big data as well as big government data, at a faster rate, thus making manual data analysis and interpretation impossible. ML does not only automate the analysis of big government data but also can provide data-driven answers. Accuracy The results of ML systems, irrespective of the used techniques, are more accurate, since ML can process big government data and no intervention from either knowledge engineers or domain experts is required. Performance & Process Simplification Easier and faster way for automated classification to analyse data when compared to manual process which would consume a significant amount of time and effort by reducing the cost. Automating E-Government Services with Artificial Intelligence: Extensive use of Machine Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021
with Artificial Intelligence: Extensive use of Machine Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021
Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021 Challenges of Machine Learning Privacy & Ethical Issues In many cases (e.g. healthcare) the collection of personal data, the ownership of personal data and the benefits of their processing leads to privacy and ethical issues. Quality & Quantity of Data Lack of data (e.g. geographical data) or even accessed data may not be representative and, in cases of predictions, barriers can be found decreasing the quality and quantity of the ML system. Heterogeneity of Data Heterogeneity of data (e.g. writing styles or different vocabularies) is a challenge for ML because it can lead to wrong results. Availability of Data Difficulties of gaining regulatory approval of accessing data (for instance in healthcare), or even lack of data (geographical data) in order for a ML system to be properly trained for quality results.
Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021 Challenges in Indonesia 1. Readiness of regulations governing ethics of use 2. Readiness of the workforce 3. Infrastructure readiness and supporting data 4. Readiness of industry and the public sector in adopting artificial intelligence innovations.
out. Discussion, any question? https:/ /ksnugroho.site https:/ /linkedin.com/in/ksnugroho https:/ /github.com/ksnugroho Automating E-Government Services with Artificial Intelligence: Extensive use of Machine Learning Kuncahyo Setyo Nugroho | Presented in Webinar of E-Government Transformation Toward Smart System 2021