The ACM Fellows Program is the organization’s most prestigious honor. It recognizes the top 1 percent of ACM members for outstanding accomplishments in computing and information technology.
“The Fellows program allows us to shine a light on landmark contributions to computing, as well as the men and women whose hard work, dedication, and inspiration are responsible for groundbreaking work that improves our lives in so many ways," ACM President Vicki L. Hanson said in announcing the new class of Fellows.
Kalé, who is the Paul and Cynthia Saylor Professor of Computer Science, was honored “for development of new parallel programming techniques and their deployment in high performance computing applications.”At Illinois, Kalé leads the Parallel Programming Laboratory, and his research focuses on a broad range of topics within parallel computing.
He led pioneering work to develop adaptive runtime systems in parallel computing and incorporated it in the Charm++ parallel programming framework. Kale and his group are known for collaborative development of several highly scalable computational science and engineering applications, including NAMD (biophysics), ChaNGa (Astronomy), and OpenAtom (Quantum Chemistry).
Kalé also is a Fellow of the Institute of Electrical and Electronics Engineers, a winner of the 2012 IEEE Computer Society Sidney Fernbach Award, and a co-winner of the 2002 Gordon Bell Prize.
Zhai, who is a Willett Faculty Scholar, was honored “for contributions to information retrieval and text data mining.”Zhai leads the Text Information Management and Analysis Group, or TIMAN. His work aims to develop new computational models and algorithms for building intelligent information systems – intelligent search engines and text analysis engines, for instance, to help people manage and exploit large amounts of data.
He and his group have contributed a large number of effective models and algorithms in information retrieval, which is the underlying science of all search engines. They have also developed many new techniques for analyzing and mining text data, including news articles, scientific literature, emails, enterprise documents, and social media, and applied them to developing novel text mining applications. Zhai has a particular interest in developing intelligent information systems to aid in medical care, education, and scientific discovery.
A recipient of the Alfred P. Sloan Research Fellowship and the Presidential Early Career Award for Scientists and Engineers, Zhai was also recognized previously as a Distinguished Scientist by Association for Computing Machinery.
ACM will formally recognize its 2017 Fellows at the annual Awards Banquet, to be held in San Francisco on June 23, 2018.
Illinois Computer Science Professors Laxmikant Kalé and ChengXiang Zhai talk about the work that went into becoming ACM Fellows, and about where they see their research heading in the future:
Q. What is the new parallel programming technique for HPC that you created?
A. I and my group have pioneered the idea of adaptive runtime systems in parallel computing, or HPC. We not only developed the ideas, but always incorporated them in a practical parallel programming system called Charm++ that we distribute and maintain from Illinois. We co-developed several science and engineering applications using Charm++, which allowed us to validate and improve the adaptive runtime techniques we were developing in our research in the context of full applications. The application codes developed include NAMD (biophysics), OpenAtom (quantum chemistry and materials modeling), ChaNGa (astronomy), Episimdemics (simulation of epidemics), and others.
Q. What HPC systems has it been deployed on, and where do you see it being deployed in the near future?
A. These are highly scalable codes that run from small clusters to supercomputers, including Blue, Waters, on hundreds of thousands of processor cores.
Q. What benefit does this bring to the larger HPC community?
A. Our approach allows parallel programmers to write code without worrying about where (on which processor) the code will execute, or which data will be on what processor.
The runtime system continuously watches the program behavior and moves data and code-execution among processors so as to automatically improve performance, via dynamic load balancing. This approach especially helps development of complex or sophisticated parallel algorithms.
Q. Becoming an ACM Fellow is obviously quite an honor. Tell me a bit about your career path and how you believe it led you to this point.
A. Apart from the novelty of automated runtime optimization, my research is characterized by continuous engagement with science applications. This is a costly path for a researcher, and I paid the price for it early in my career. But I believed such application-oriented yet computer science-centered research is the only way to try to make a lasting impact.
The credit for my success and for this award certainly goes to generations of my students who worked on various aspects of adaptive runtime systems.
Q. What does it mean for you, personally and from a career perspective, to become an ACM Fellow?
A. I'm really honored to be recognized as an ACM Fellow. It's a recognition of the impact of my research, which is always rewarding for a researcher since our goal is to discover new knowledge and invent new technology that can have substantial effects on our society.
My contributions are mostly made through joint publications with many students and collaborators, this title is also a recognition of their efforts. This recognition also sets a higher standard for my future research.
Q. Your work on information retrieval models appears to be aimed toward an ultimate goal, but it seems like there are many smaller-but-important goals along the way, a number of which you’ve already met. Can you talk a bit about how those achievements build on each other?
A. A modern search engine such as Google or Bing is a very complex system with many components to be optimized, including how to interpret a user's query intent, how to understand the content of documents, how to accurately match a query with each document, how to present search results to users, and how to engage users in a collaborative way to help users finish a search task with minimum effort. In my research, we have developed new techniques for improving many of those components that in principle can be combined to achieve an additive benefit.
Q. What makes the goal you mention of finding one information retrieval model so challenging?
A. Among all the components, the retrieval model used by a search engine is probably most important because it directly determines the accuracy, and thus also the usefulness of the ranked list a search engine returned to a user. Because a user's query can be on an arbitrary topic and a relevant document doesn't always use the same words as those in the query, developing an effective model that works well for all the queries is very difficult.
Q. Where do you see your work being deployed in the future?
A. While many algorithms we developed are general and can be applied to many different application domains, I am particularly interested in applying them to improve health care, improve online education, and accelerate scientific discovery. A general goal is to develop intelligent software assistants that can help humans optimize decision making.
In the healthcare domain, I'm currently working on analysis of electronic medical records to enable precision medicine. In the education domain, I'm working towards intelligent interactive tutoring agents and building a virtual data science lab to enable students to experiment with big-data algorithms on the cloud. I'm also working toward building an intelligent researchers' workbench to help researchers improve productivity, accelerating scientific discovery.