- This event has passed.
ICADS ’23: Second International Conference on Applied Data Science
July 25 @ 8:00 am - 12:30 pm
…Smart Classroom and Smart Home Applications, Machine Learning Tools in the Cloud, …
Panel discussion “Should AI development really take a pause?”
8am – 9am PT AI-based Decision Frameworks for Smart Environments – Some Case Studies, Prof. Ram Mohana Reddy Guddeti, India
9am – 10am PT Performance-based content generation for language learning, Dr. Kostas Karpouzis, Panteion University of Social and Political Sciences in Athens, Greece.
10am – 11am PT Panel Discussion: Should AI development really take a pause?
Dr. Patrikakis Charalampos, University of West Attica, Greece
Dr. Vishnu S. Pendyala, San Jose State University, CA, USA
Manish Mradul, Director, Palo Alto Networks
Moderator: Krupa Kothadia
11am – 12pm PT Using Machine Learning Tools in the Cloud: Experience Gained from the Ask4Summary research project
Prof. Maiga Chang, Athabasca University, Canada
AI-based Decision Frameworks for Smart Environments – Some Case Studies
Online Examination Question Classification Systems: The advancement in education has emphasized the need to develop an efficient online examination (exam in short) system where we carry out the assessment of the exam questions qualitatively and the cognitive levels of students mapped to the Bloom’s cognitive levels; thus, evaluating subject-related learning by the students. Hence, a novel optimized framework, referred to as QC-DcCapsGAN-AOSA, is developed by combining the Dual-channel Capsule Generative Adversarial Network (DcCapsGAN) with the Atomic Orbital Search Algorithm (AOSA) for preprocessing a real-time online exam questions dataset of Indian universities, thus identifying the key features from the raw data using Term Frequency Inverse Document Frequency (TF-IDF) and finally classifying the exam questions.
We train on this dataset using a federated learning approach to ensure the privacy of the students. The performance of both the centralized and decentralized (federated) models with transfer learning is carried out in terms of accuracy, recall, precision and F1-Score metrics. The results demonstrate that the federated learning with DenseNet121 transfer learning model is superior to all other considered models.
I will also discuss Smart Home Kitchen Appliance, Refrigerator and Gas Leakage Prediction and Alert System.
Performance-based content generation for language learning
Generation of appropriate content for serious/educational games is an extremely important concept since it can make all the difference between adoption and retainment of the game, which increase the possibility to achieve its learning objectives and attrition. In the iRead project, we are creating a serious game and supporting applications for entry-level language learning. It is a flexible way to model language, teaching priorities and student mastery, allowing for personalized learning and student analytics.
This modelling approach can be extended to different languages, as long as each of them can be modelled as a set of features, taught sequentially; based on that, personalized content selection may be used to select words, sentences or passages of text, suited for each student based on their performance and teacher-selected learning objectives.
Panel Discussion: Should AI development really take a pause?
Recently, the news media has been filled with reports and debates on the powers and dangers of Artificial Intelligence. Adding to the confusion, even experts have been differing in their opinions. This panel discussion is intended to help clarify and provide interesting perspectives on the topic.
Using Machine Learning Tools in the Cloud: Experience Gained from the Ask4Summary research project
Ask4Summary (https://ask4summary.vipresearch.ca/) is a research that has a system periodically running backend services to process text-based content (e.g., a course’s learning materials and the CORD-19 dataset). The CORD-19 dataset includes a growing number of academic articles regarding Coronaviruses; at present, there are more than 717,000 full text articles in the CORD-19 dataset. When Ask4Summary processes these text-based content, it uses Natural Language Processing (NLP) techniques that include tokenization, n-grams extraction, and part-of-speech (PoS) tagging. It then identifies the keywords from a user’s question and uses cosine similarity to summarize the associated content. Currently Ask4Summary can serve users via three ways: web system, Moodle plugin, and chatbot. In this talk, I would also like to share some of our experiences and insights from using Amazon’s Comprehend Keyphrase Extraction and Syntax Analysis APIs in the Ask4Summary research project.
Speaker bios and regitration (Free):