Andrew Thangaraj

Title: Learning from IITM's Data Science Program

Abstract: : IIT Madras' undergraduate data science program (onlinedegree.iitm.ac.in), started in Jan 2021, with online course delivery and in-person exams, has more than 15,000 enrolled students today. How can high quality education in data science be delivered to such a large number of students? How is quality measured and maintained? How are the students really doing? The talk will touch upon answers to the above questions and further elaborate on the promise such programs hold for the future.

Sriram Rajamani

Title: AI Assisted Programming

Abstract: : We present a vision for a futuristic programming environment, where ML models and logical rules co-exist and evolve over time. We substantiate our vision using three case studies: (1) safe code generation using large language models, (2) Heterogeneous data extraction, and (2) regulatory compliance for online advertising. Using these case studies, we hypothesize what such a futuristic programming system can do to balance productivity with safety, security and reliability, present new research results, and point to directions for future work.

N S Kumar

Title: Teaching Data Structures - My View

Abstract: : Data structure remains to be a core course in the study for a degree in Computer Science. Is this being taught the right way? More often than not, a data structure is taught as an implementation than an interface. A list is always considered as a linked list. Alternate implementations for the interface of a data structure are rarely discussed. Separation of interface from implementation leading to flexibility in changing the implementation without affecting the client is not highlighted. Any data structure would have number of components. Could we refer to a component without exposing the implementation? Could we have an opaque structure standing for the position? Could we have a way to traverse the data structure if it is semantically allowed? Do all data structure libraries support this concept? If yes, why is that we do not teach this concept in our courses? This talk shall try to address the following features:

1. Highlight the concepts of interface and implementation
2. Provide an opaque data structure for locating a component and traversing the data structure
3. Show that an interface can remain immutable when the implementation is changed

Brett Becker

Title: It's a Two-Way Street: From the Computing Classroom to Computing Education Research

Abstract: There is much discussion in computing education about turning research into practice because research-informed teaching could (or should) be more effective. However, it is often noted that there is a disconnect between practice and research. A lot of research rarely informs classroom practice, and often classroom practice is not informed by research. Nonetheless, for research to inform practice we need research. Where does this research come from? How does it become research? In addition to being informed by research, classroom practice can be one source of research, and the classroom itself can be the laboratory. The result is a two-way street between practice and research. While some research - for instance the presentation of some theories, literature reviews, and research not directly involving students - may not originate in the classroom, much research does originate in, or takes place in, the classroom. This talk describes several routes from the classroom to research. It will cover topics such as scoping classroom-based research studies, ethical review, approaches and methods, data collection and analysis, the role of curriculum, research presentation and paper types, venue choice, the submission and review process, publication, and post-publication considerations. The goals are to help those that are less familiar with computing education research to conduct research in their classrooms and additionally to help those in the classroom use research in their practice.

Yogesh Simmhan

Title: Pedagogic Practices for Teaching Distributed Systems Courses

Abstract: : Parallel and Distributed systems such as cluster computing, cloud computing and Internet of Things (IoT) have become pervasive technologies that are essential not just to computer science but also to design and deploy applications from interdisciplinary and data science domains. Under-graduate and post-graduate curriculum needs to incorporate these as essential or elective courses to ensure that students are well trained for their careers. This requires imparting systems concepts, teaching programming models and conducting hands-on labs to translate theory to practise. This talk will discuss experiences and best practices in teaching systems courses and designing systems labs, and briefly review a model curriculum for distributed and cloud systems micro-specialization.

Hemangee K. Kapoor

Title: DEI activities at the ACM and How to make CS education more inclusive

Abstract: : Research has demonstrated that diverse teams are more innovative and productive. At the ACM, as a computing professional community, it is our responsibility to make the resources and services cater to the needs of all members of the society. To ensure that the governance and activities of ACM involve members from diverse background and perspectives, the Diversity, Equity and Inclusion (DEI) council has been established. The talk will begin by describing the charter of the DEI council and its main responsibilities. We would then discuss how the responsibilities can be achieved by making three broad categories. The various initiatives that are planned to be executed in each category will be discussed; followed by the ongoing activities. Case studies/events/contributions/ done in order to bring DEI in our daily professional activities will be highlighted. These will include achievements and initiatives at ACM global as well as ACM India level. Subsequently, we shall have an open discussion on how to inculcate DEI in CSE education.

Smruti R. Sarangi

(OCCE Award Winner 2022)

Title: Perspectives on Teaching Computer Architecture in Developing Countries

Abstract: : Different areas of computer science have different challenges with respect to curriculum development, textbook writing, and teaching. Computer architecture represents one extrema in this hyper-dimensional space: it is an applied discipline, the theory for many sub-areas is still in development, and unlike traditional theoretical areas, problem-solving and tutorial based teaching is still uncommon. Students typically appreciate the rigor of traditional algorithm courses because they are similar to the courses they have seen in high school, however computer architecture is an engineering discipline: it is a science, an art, and sometimes a combination of both. As the name suggests, it is quite like traditional "architecture" (designing beautiful buildings ... ). To further complicate matters, most textbooks in this subject have originally been written in the late nineties and fail to capture most of the developments that have happened over the last two decades, which are fundamental and revolutionary at the same time. In the late nineties the field was an applied discipline where the design of systems, the specific design choices made by engineers, and a description of important architectures was considered proper pedagogical practice. This was probably acceptable in developed countries where students had an adequate exposure to computers since their teenage years. However, this approach did not work in India and a few other developing countries where I taught. Hence, there was a need to grow the field and instead focus on providing timeless insights that are not dependent on specific architectures or specific technologies. There was a need to bring the field in line with other fields such as chemistry, physics, biology, and of course traditional CS (data structures, algorithms, discrete math, etc.). This was the nature of the challenge that I faced when I started writing my two books on computer architecture (www.basiccomparch.com and www.advcomparch.com). There was a need to formalize basic notions, provide a theoretical framework, and create a discipline that is more in line with traditional theoretical subjects, which the students (at least in developing countries) are more used to. Hence, a novel pedagogical approach was developed where the connect between Turing machines and basic architectures was shown. The same approach was followed to explain all the major concepts by relating them to other other concepts in traditional math or CS: caches were connected with hashtables, virtual memory was connected with indexing, I/O was connected with networking, OOO processing was connected with graph algorithms, and memory models were re-explained with novel concepts developed by the verification community. The end result of doing so was 1500 pages of computer architecture (captured in my two books). Many concepts appear to be intuitive such as memory consistency and routing algorithms in on-chip networks. However, this never leads to a deeper understanding and cannot lead to practical designs because there is not enough knowledge to write real code and capture corner cases. This is where formal training and a connect with well-established CS concepts is required. In my second book on Advanced Computer Architecture, especially in the later half, I focused on power management, security, reliability, and architectures for AI/ML. Here again, the theory was not well developed and thus there were no popular mathematical formalisms. New notations and new theoretical tools were introduced to put these areas on a sound footing. This is by far the most important contribution of my two books, which is to formalize and "mathematicalize" concepts in computer architecture that hitherto had very applied definitions that were often prone to misinterpretations. Towards the end of the talk, we will also digress into some of my other academic adventures such as co-chairing a committee to create the CBSE CS curriculum for schools, a lab course on processor design and kernel hacking (future of OS teaching), and my take on the future of teaching and learning (in general).