People of ACM - Meenakshi D’Souza

October 10, 2024

What made you decide to pursue a PhD in computer science? What was the landscape like for women in India to get jobs in academiaand industry when you graduated?

I graduated with a Gold Medal and a Master's Degree in Mathematics, and a specialization in certain aspects of theoretical computer science. I wanted to pursue higher studies in computer science and a few faculty members from my alma mater encouraged me to apply for a PhD. I was lucky to secure admission in three institutions and chose to pursue a PhD in Theoretical Computer Science from the Institute of Mathematical Sciences in my hometown of Chennai.

In the late 1990s, Mathematics graduates could easily secure campus placements in the banking, financial, and IT sectors. An additional degree in Education (Pedagogy) was mandatory to get a teaching position in a good high school, a popular choice for young women graduates then. Since I enjoyed learning subjects like Graph Theory, Automata Theory, and Programming, I choose to pursue higher studies in Theoretical Computer Science.

Your most downloaded paper in the ACM Digital Library is “Improving Estimation Accuracy by Using Case-Based Reasoning and a Combined Estimation Approach.” Along with co-author Srinivasa Gopal, you examined the different models for estimating software for estimates related to software project cost, defects, personnel, required, etc. What insight did you offer in this paper to improve software estimation models?

This was an initial work on software defect estimation that I had gained experience in during my tenure at Honeywell Technology Solutions Lab, Bangalore. While I was consulting for a few firms working on software quality, I happened to notice that many of them wanted to estimate several metrics for software quality and improve upon their practices through accurate estimation. This piece of work proposed an algorithmic approach based on case-based reasoning for improving software quality estimates, and subsequently has been implemented in two firms that work on software quality and testing.

What is an example of an interesting research challenge you are working on now at IIIT Bangalore that you think will have a significant impact on your field?

At IIIT-Bangalore, I have been working in the areas of software testing and formal methods, focused on embedded, safety-critical software. We have worked on collaborative projects with MathWorks, ABB, Samsung, and Siemens, and with several Indian government organizations including the Department of Science and Technology, the Defense Research and Development Organization, and more recently, the Aeronautical Development Agency. Application domains include robotics, programmable logic controllers, push-to-talk systems, real-time operating systems, Internet of Things (IoT), and applications that (will) include AI and ML algorithms.

Ensuring that this safety-critical software is nearly free of errors is important, and even exhaustive testing cannot guarantee the absence of errors. Many of these applications implement complex algorithms and are built over real-time platforms that are themselves difficult to model and verify. Our work involves both formal verification and testing, as well as developing suitable abstractions and algorithms that will guarantee certain correctness properties of these software. One recent research challenge concerns the safety and robustness of the AI and ML algorithms that will be used in safety-critical applications for specific tasks like situational awareness, perception, predictive monitoring etc. The underlying models and learning architectures are inherently difficult to test and verify, and when used in tandem with deterministic software, they add to the complexity of the application. We are currently working on testing algorithms for applications which use multiple neural networks, and also on algorithms and guidelines for independent verification and validation of these applications.

In a recent video, you noted that women are well represented in undergraduate STEM programs at Indian Universities (43% of STEM graduates are women). However, you also note that only 14% of these women graduates take up careers in STEM. Why does India do a better job than other countries in enrolling and retaining women in undergraduate STEM programs? Why don’t more Indian women pursue careers in the field?

Women are well represented in under-graduate STEM programs in Indian universities thanks to the growing urbanization of education, preference for studying science by an increasing number of students, and other socio-cultural factors. There are a good number of women pursuing education in the core sciences, mathematics, and also medicine, but women are heavily underrepresented in engineering education, specifically in India's top engineering institutes such as the Indian Institutes of Technology (IITs). I would like to point to a November 2022 Communications of the ACM article, where we indicate that the pipeline of high school girls students studying science who later pursue Engineering and take up jobs is very leaky. The percentage of women studying engineering is low, motivating the IITs (and subsequently several other institutions) to implement a super-numerary quota-based reservation of seats exclusively to women students. This has been a successful effort, increasing the percentage of women in India's top engineering institutions to 20%.

The employability of Engineering graduates in India is low in general and women face additional challenges in terms of restricted geographical mobility, pressure to get married and take up full-time family responsibilities, as well as lack of access to upskilling and re-skilling. This has resulted in many women opting to not pursue higher education, dropping out of their jobs by the time they gain 5+ years of experience, preventing them from ascending the career ladder or taking up research.

As President, what are some of your goals for the ACM India Council? How will Indian computing professionals benefit by joining this organization?

The activities of the ACM India Council involve both academia and industry. Our efforts include supporting research through events and travel grants exclusively for research scholars, organization of high-quality computing conferences, advancing computer science education in high school and college levels, and awards to recognize teaching and research achievements. ACM-W India has an exclusive focus on activities that lead to empowerment of women in computing which is close to my heart.

ACM India membership is high among aspiring private educational institutions and universities across India. The student and professional chapters in these institutions pursue several different activities. As President, I would like to expand our activities to augment the local, chapter-based activities so that they lead to an overall improvement in teaching and learning in these institutions. We on the ACM India Council are also working towards increasing the presence and the scope of ACM activities in India's elite engineering institutions. This, we believe, will in turn help ACM gain visibility amongst the young computing and IT professionals of India.