Top 2023 IEEE Innovation In Societal Infrastructure Awardee Kathleen McKeown Shares Deep Insights

Top 2023 IEEE Innovation In Societal Infrastructure Awardee Kathleen McKeown Shares Deep Insights

Professor Kathleen R. McKeown is the recipient of the 2023 IEEE Innovation in Societal Infrastructure Award—the highest award recognizing significant technological achievements and contributions to the establishment, development, and proliferation of innovative societal infrastructure systems through the application of information technology with an emphasis on distributed computing systems.

Professor Kathleen R. McKeown’s remarkably compelling research, global societal impact, deep insights, lessons, innovations, and exciting narratives of discovery are explored in this extensive interview which is unscripted and provided in full below.

The IEEE, Institute of Electrical and Electronic Engineers, its roots dating back to 1884, and with more than 420,000 members in 160-plus countries, is the world’s largest technical professional organization dedicated to advancing technology for the benefit of humanity. Professor Kathleen R. McKeown embodies all of the excellence in this iconic organization.

This article is based upon insights from my daily pro bono work, across more than 100 global projects and communities, with more than 400,000 CEOs, investors, scientists, and notable experts.

Professor Kathleen R. McKeown’s Brief Profile

Professor McKeown’s profile is so extensive that a summary is provided with the non-profit IEEE TEMS (see interview series – Stephen Ibaraki – “Transformational Leadership and Innovation…”). This direct link to the interview page contains the profile and video interview.

Here’s the profile summary.

Kathleen R. McKeown is the Henry and Gertrude Rothschild Professor of Computer Science at Columbia University and the Founding Director of the Data Science Institute, serving as Director from 2012 to 2017. She is also an Amazon Scholar. In the past she served as Department Chair (1998-2003) and as Vice Dean for Research for the School of Engineering and Applied Science (2010-2012). McKeown received her PhD from the University of Pennsylvania in Computer and Information Sciences in 1982.

A leading scholar and researcher in the field of natural language processing, McKeown focuses her research on the use of data for societal problems. Her research has shown how to learn from knowledge of past disasters, as seen through the lens of the media, in order to provide updates on disasters as they unfold. This resulted in a system to automatically generating update summaries about disasters. Her work also provided techniques for understanding subjective, personal narratives of those who have experienced disaster, enabling automatic identification of the most reportable event, or climax, of the story. Finally, her team developed approaches that use sentiment detection over social media in low resource languages to identify where people are still suffering after a disaster.

In current work on social media analysis, an interdisciplinary group she leads, involving social work, linguistics and psychiatry, is analyzing posts from Black communities to identify expressions of online grief in response to the events of our times. The project, Identifying and Understanding Digital Expressions of Black Grief, will develop a system to automatically identify digital expressions of grief that could then be used by social workers and health professionals for intervention and treatment programs. The project is using insight from social work to analyze posts to identify the triggers of expressed emotion. A linguist on the team with experience in African American English is analyzing posts to annotate how emotions are expressed. McKeown’s group is developing a computational model specifically tuned for African American English to identify expressions of grief and events causing the emotion.

In earlier years, she led a team on the development of a system for personalized search and summarization over medical literature, allowing clinicians to find information relevant to their patients and patients to find information relevant to their situation. By making use of patient characteristics from the patient record, including information about diagnoses, the search component was able to retrieve journal articles that were a better match for the patient situation and the summarization component could highlight information in the article that was relevant.

Over her career, McKeown has also focused on supporting women in the field. She was the first woman to receive tenure at Columbia’s School of Engineering and Applied Science in 1989. She later led Columbia’s Commission on the Status of Women to develop a parental leave policy for the university, which was adopted by the University Senate in the mid-nineties. She supported the establishment of Columbia’s Women in Computer Science group when she served as Department Chair; this group is now a thriving community of more than 100 undergraduate and graduate students. She strives for strong representation of women in her PhD group; roughly 45% of her graduated PhDs have been women, more than double the national average.

She has received numerous honors and awards, including American Academy of Arts and Science elected member, American Philosophical Society elected member, American Association of Artificial Intelligence Fellow, a Founding Fellow of the Association for Computational Linguistics and an Association for Computing Machinery Fellow. Early on she received the National Science Foundation Presidential Young Investigator Award, and a National Science Foundation Faculty Award for Women. In 2010, she won both the Columbia Great Teacher Award—an honor bestowed by the students—and the Anita Borg Woman of Vision Award for Innovation for her work on summarization of news. Her most recent honor is the 2023 IEEE Innovation in Societal Infrastructure Award.

Interview with Professor Kathleen R. McKeown

AI is employed to generate the transcript which is then edited for brevity, clarity while staying with the cadence of the chat. AI has an approximate 80% accuracy so going to the full engaging video interview is recommended for full precision. Time stamps are provided however with the caveat that they are approximate.

The interview is recommended for all audiences from students to global leaders in government, industry, investments, NGOs, United Nations, scientific and technical organizations, academia, education, media, translational research and development, interdisciplinary and multidisciplinary work and much more.

Stephen Ibaraki 00:00

Hey Kathy, you have such an outstanding record of contribution across so many different domains and interdisciplinary. You just received this amazing IEEE award for really a lifetime accomplishment, but you have so many other recognitions. Even things like your students, selecting you as a top teacher. I mean, it is just an incredible background. I really appreciate you coming in and sharing your insights today.

Kathleen R. McKeown 00:35

Well, thank you so much for having me. I’m looking forward to our conversation.

I wouldn’t be here where I am today, if it wasn’t for my students who have really been amazing to work with.

Stephen Ibaraki 00:52

You definitely exhibit this idea of diversity, equity and inclusion, but for your entire career, and my audience is really diverse. They’re scientists, notable experts, CEOs and investors. It’s quite broad. So I’m always curious and my audiences, as well. What really contributed to making this wonderful scientist and contributor that you are today. Kathy, are there one, two, maybe three inflection points? Perhaps it was really early in life or through a mentor, or it could be a series of things? What created this outstanding person?

Kathleen R. McKeown 01:32

I think it’s really a series of things. I would have to first go back to my parents who were really amazing. They were both scientists. My mother was an applied mathematician who worked throughout her life, through the 60s, when I was growing up. This was a time when women didn’t work. I had an amazing role model in my mother.

I was also very strongly influenced by my advisor when I started my PhD. As well as other women in the field, who were very prominent. I think, in particular, of Karen Spärck Jones (pioneering British computer scientist), who mentored me, in the area of summarization. She was a professor at Cambridge University.

And then I think, maybe over time, working with my students, and wanting to to do things that made them happy, made them feel valued in the world. I think in particular, when Hurricane Sandy hit New York, this was a time when students wanted to give back. It started our work in/on disaster—on developing computational systems to help with disaster.

Stephen Ibaraki 01:36

That’s really an interesting confluence. From your family, to mentors you’ve had, to your students. Just a broad intermix of enabling you to do the work you do and continue to do.

Let’s talk about some of the work that you’ve done in the past. Can you talk about your work in natural language processing? You’re really renowned for your focus on words and the use of data for societal problems. Can you talk more about that?

Kathleen R. McKeown 03:55

Oh, yes. My interest in natural language processing sort of came together as a result of two interests that I had as an undergraduate. I couldn’t decide whether to major in comparative literature or in mathematics. I didn’t know about computer science at that time. I chose comparative literature, but I discovered computer science at the end of my undergraduate. Following my undergraduate was when I learned about natural language processing, which would allow me to bring my two interests together. That was when I applied for a PhD in Computer Science at the University of Pennsylvania where there was a strong focus on natural language processing. I was very thrilled. I guess this was early 2000s, when we did some joint work with faculty from the English Department on analyzing novels. Identifying social networks within novels. So it sort of brought my two interests, full circle back to my roots, essentially.

Stephen Ibaraki 05:31

There’s also how you applied your knowledge of past disasters. There’s research about this, seeing the lens of media, and provide updates on disasters as they unfold. Can you talk more about that work?

Kathleen R. McKeown 05:46

Sure. As I mentioned, I was motivated to begin this work, when Hurricane Sandy came through and hit New York. Many of the students were really interested in thinking about how they could give back to the community. Some of them had been personally impacted. At that point in time, we had already developed a system called (Columbia’s) Newsblaster, which could summarize news on a daily basis where essentially, we are similar in many ways to Google News. That was very early times. We had developed an approach where it could identify events of the day. Multiple articles that were reporting on those events and it could generate a summary that synthesized information from multiple articles on a single event. We had looked at a number; we used this as a platform for research. So different kinds of research that would take place there. One thing that seemed particularly interesting; to be able to take an event, which was occurring at that point in time and looking at how it unfolds over time. Each day, to be able to give an update on what had happened over the day before. We looked at this within the context of the disaster. To be able to provide updates from news that was coming in. On what was new today. What did we know today? Or even in that system, we were looking, hour by hour. What did we know in this hour that we didn’t know, in the hour before? As it happened there, there was some interest internationally in this topic at that point in time. There was a competition essentially, for researchers who were working in this area. And so, we participated in that.

Stephen Ibaraki 08:24

I do quite a bit of work with different UN agencies. Some of them are bigger than what we think of the UN itself; the one out of New York. Bigger budgets. The World Food Programme, for example. They have something like 100 aircraft, 30 ships, 6000 trucks, maybe about 20 to 24,000 staff globally. They would make use of this kind of signaling system. Have you worked with the United Nations agencies like the World Food Programme, or others?

Kathleen R. McKeown 08:56

We have not. We did this as a research project. Our (Columbia’s) Newsblaster system, we did have up and running on the web for years. We did incorporate in that the ability to provide update summaries, but not the ones specifically about disaster. I think that could be a good kind of thing going forward. It requires a certain amount of work that is not research in order to deploy a system. And so I think given that I’m working with PhD students, this is why we stopped at this point in time.

Stephen Ibaraki 09:50

I guess you’re really talking about translational research that is deployable as well. The (interview) audience also consists of entrepreneurs. As they look at the body of work and deploy them as global solutions.

I can see this then being founded on your earlier work where you led a team on the development of a system for personalized search and summarization over medical literature. Can you talk about that work?

Kathleen R. McKeown 10:23

Sure, that was an interdisciplinary project. It was between computer science and biomedical informatics. As well as physicians in the medical school at Columbia. There we were, particularly interested in how we could personalize and make available information on a personalized basis to both physicians and patients and their families.

On the physician side, we thought, we have one of the hardest things when you’re developing a personalized system is, how do you get information about the end user? If we want to personalize for the physician, information that the physician is seeking to help and treatment of their patients? How do we personalize that? Tailor it for the patients that the physician is treating?

We thought, well, we have at the medical school, the patient record. Columbia was one of the earlier leaders in establishing an online patient record. We actually have a large amount of information about the patient, their characteristics, their diagnosis, their course of treatment. We could use those characteristics to help us to find journal articles that were more appropriate for the particular patient under the physician care.

That’s sort of how we started off. Having a team where we could draw on the expertise of physicians in the hospital. We could work to design the system that better met their needs.

We also, at that point in time, looked at a developing a similar approach for patients where we would want to be able to provide information that was accessible to the lay population. And again, that was especially targeted towards their needs at that point in time.

Stephen Ibaraki 13:06

I can just see the utility of this across all the different sort of domains that I’m in. So really foundational work.

I know, in your current work on social media analysis, you work with interdisciplinary researchers. It’s really fascinating and involves social work, linguistics, and psychiatry. You’re analyzing posts about the black community. Can you talk more about this project of identifying and understanding digital expressions of grief?

Kathleen R. McKeown 13:41

Yes. So, again, we’re very motivated by the events of our time. I had been working with my collaborator, the faculty member from social work. His name is Desmond Patton (Professor at the Columbia School of Social Work and Department of Sociology), for multiple years before we started this project.

But we were particularly concerned with how the black community has been hit; they are traumatized by events both with COVID or with violence, or police … with problems along those lines.

And so, when we worked together before, we have a very nice sort of collaboration. The Social Work team is very interested in doing what they call, the qualitative analysis of online social media posts. We’re looking here actually at two ways of gathering those posts. One, we’re looking at online forums. But we’re also have developed our own platform where people can write about the experiences that they have been through.

The Social Work team does this qualitative analysis, which is to be able to identify the different emotions that people are experiencing, and the events that have triggered those emotions. Given their background and familiarity with the community, they can provide a contextual understanding of why these people may be experiencing what they are.

We also have a linguist on the team and her expertise. Her name is Jessi Grieser. She has just moved to University of Michigan. Her expertise is in African American English. Looking at the specific linguistic expressions that are used.

We also have a psychiatrist on the team, Kathy Shear who’s at Columbia. Her expertise is in grief, the different stages of grief and how people experience it.

So, given this group, we have the potential to create a corpus of expressions. Posts that people have written with layered annotations, which represent the meanings, the context in which people have had the experience. The particulars of the linguistic expressions that they use to express their grief, or their stance on particular topics.

And then Kathy Shear will do interviews with people, so that we get who have posted; so that we get more background about why they were feeling the way they were doing. This gives us a very rich representation, from which we can learn how to develop a system that can automatically detect when people are experiencing grief or have gone through traumatic events.

One thing that I would note about our model is that we are focusing on developing a computational model of African American English. This has been something that I think has been lacking within the natural language community, where the focus is on very large language models, which draw on mixed communities on the web, that are primarily I would say, standard American English.

Stephen Ibaraki 18:09

That’s fascinating work. Especially when you hear about … I could see the implications of what you’re doing on the Diversity Equity Inclusion side.

There are these very large (AI/ML) language models: GODEL from Microsoft; LaMDA from Google; GPT-3 from the OpenAI group and so on, or DALL-E-2 (OpenAI) applied to the art arena. What are your thoughts on that, as these models become much bigger and bigger … using diffusion techniques, or transformers and so on? Your views on that, and then the implications of your work with these very large models that are out there?

Kathleen R. McKeown 18:54

Well, there’s no doubt that these very large language models have been quite successful. We see dramatic improvements in a lot of the tasks that we address in the natural language community. So, summarization, which is clearly the area that I work in has seen tremendous progress with the use of these large language models.

But I think people are discovering problems with bias in these language models. They’re sometimes hard to control. It’s possible that they produce output that is unintended. For example, in their use in summarization systems, we sometimes see that the summary contains information which did not occur in the input article; you know, sort of appeared there. People are working on that problem. I have no doubt that there will be progress in addressing them.

But the problem of bias and developing models that are tuned for specific populations … (pause for dog wanting attention) … yes, the problem of bias is really something that has come out more recently and that people are looking at more seriously. It’s a difficult problem. Because, when these language models are trained on huge amounts of data, and to get that kind of data, we are often gathering texts that date back years. Stereotypes are embedded in that text. It can be hard to tease them out. I will also think of my collaborator Desmond pointed out, in an article at one point in time, which was somewhat eye opening. Which talked about the impact on people of color who communicate with the dialogue systems? And they don’t speak their language. How does that feel to have to communicate, when you want to make use of technology, to have to change your language to be able to be understood? How does that make you feel, as far as being included in a community? I think it is very important that we develop language models that can understand and even ultimately generate in dialects that are not standard American English.

Advertisement

Stephen Ibaraki 22:36

Again, because there’s just so much work in this. There’s an attempt that the large language models can get more generalized in their capability. Do you think that’s going to happen? So, it’s not narrow. It’s this whole aspect of AGI or Artificial General Intelligence? Do you see these models moving in that direction?

And then if we can eliminate the bias or reduce the risks associated with bias and where it becomes more generalized in its applications to various areas? Or do you think it’s not going to go there? What are your thoughts on that?

Kathleen R. McKeown 23:17

I think people are definitely looking at it. I think it’s an important topic right now.

There’s a lot of focus on bias and ethics in the AI community. It is more recent. I think it takes some time after a model comes out to realize, what the problems may be? You need to use it in over a period of time.

These technologies are evolving very rapidly. It’s very rapid. So, being able to discover what has happened; it has to happen in a very short period of time. I think there will be focus. There is focus on how to address these ethical questions with large language models. How it evolves? I’m not sure.

Yes, we have multiple models that you use in different situations where there are approaches which develop the sort of mitigations that are inserted. So, you have one general model and there are mitigating techniques that are within it. These are questions and we have yet to see how it will evolve.

Stephen Ibaraki 24:46

If I can get you to look into your crystal ball of where things are going to evolve in your research. Where do you see that trending? What are some questions that you have in your mind? And that you want to put some attention to next year or 2025? Or, afterwards that you think; that’s an interesting area; I really want to focus some energy in these areas. Can you talk about it now? Or not?

Kathleen R. McKeown 25:16

Oh, yes, I can talk about it. We have some projects that were very early on, which are actually related to some of what we’ve been talking about now, but a little different spin.

So, one project, which we are starting on now, is looking at cross cultural understanding. We’re looking at, when you have communication between people who come from two different cultural backgrounds, who speak different languages. When does communications breakdown? How can we detect when people are not really understanding each other? Perhaps, because there they use different communication norms. Can we follow a conversation, and between people from two different cultural backgrounds and identify when the communication is heading towards failure, or when it’s heading towards success?

I haven’t talked about this here. But I have done a fair amount of work in multilingual research, where we do work in different languages. And again, that’s important in the global world to help people from different cultural backgrounds, who speak different languages, be able to communicate with each other. So that is one direction.

Another direction, which I think is important, we’ve been looking at misinformation from different points of view. Obviously, a goal is to be able to identify when misinformation appears. We have done some work on fact checking, for example, with information about COVID. That might appear in the news. But we’re also looking at sort of whether we can identify intentions behind misinformation. For example, whether it was accidental, or whether it was intentional, for different kinds of reasons. That’s another topic.

I am also really interested in; there’s a little bit of this, in the IEEE Award, on analyzing personal narrative. I mentioned about our work on identifying social networks within novels. We have done recent work on summarization of novels. And this is quite hard. In fact, I think it can be hard even for people to commit to that area, because it’s such a difficult topic. But I really like working on difficult topics. I am hopeful that I will have some students who will be interested in working on this with me going forward.

Stephen Ibaraki 28:52

That’s really, really fascinating, all these different elements that you’re looking at. I can see it touch points to some of the work that I do in the business area. So, across CEOs and the UN and so on.

Are there any outlier kind of topic areas that you’re thinking about? Maybe not so mainstream, to the work you’re doing, that you think are really interesting?

You’re curious about doing some additional reading in the area, though, your research may not touch on it? It could be things like consciousness, or climate. That’s why I call it, not within the scope of your work, but on the fringes or on the outlier areas, where you’re curious and follow some of that work. Or, supercomputing where exascale (billion billion operations per second) is already out, or quantum computing …?

Kathleen R. McKeown 29:54

I’m a very focused person. I realize that over the years, If we look at what I’ve worked on, from when I started my PhD in 1982, which was on generating text, paragraph length text, something that people weren’t doing at that point in time, to where I am now, generating paragraph length summaries. You can see the theme has remained throughout the years.

Where I’ve go on in directions that don’t fall into that; have been more in analysis of social media, which I can imagine, that also doesn’t answer your question, because it also has to do with natural language processing.

There is some interesting work now, which I’ve begun to become aware of, and people at Columbia are working in this area on studying the brain, and how functions in the brain correspond to how we produce language. And sort of the intersection between those two fields. I know very little about it at this point. But that’s an area that looks interesting to me.

I have worked in the past in areas of sustainable energy. So with people in the climate groups at Columbia are looking at how we can change behavior of people in regard to conserving energy through messaging. I feel that was a hard area for us to do research in. It’s hard to have the kind of data that we wanted to be able to do something really novel. So yes, there are some areas but I’m very focused.

Stephen Ibaraki 32:13

Clearly and making tremendous contributions as a result.

Kathy, you’ve done a lot of work as a pioneer, building ecosystems for women, and supporting women. Let’s get through that area. I mean, you were the first woman to receive tenure, at Columbia School of Engineering and Applied Science in 1989. What was that like? What were the challenges; how did you master those challenges at that time?

Kathleen R. McKeown 32:47

Well, it was a different world when I first joined the School of Engineering and Applied Science. There was another woman who was there for a while, but she was not there the whole time.

There were quite a few periods of time when I would go to faculty meetings for the school, and it would be a sea of men, black suits, and then me. I definitely felt different in that time …

The first other woman to join the department which happened, maybe three years or so after I joined, made a tremendous difference to me. Just having one other woman in the room I felt helped a lot.

Noticing that … that makes me notice what could help undergraduates and graduate students as they go through. Certainly, having a critical mass of women; of having somebody else who you can talk to. There is this theory that when you’re in a meeting, and a woman says something that may be ignored, and later on, a man will pick up the same topic and repeat it, and it will be responded to. I certainly have experienced that. But I think when you have multiple women in the room, that’s less likely to happen. I think for undergraduates, in particular, you really need to have a cohort and people who think like you, who you can go to and who you can talk with and who you can be open with. And so yes, I have worked to support them.

Stephen Ibaraki 34:56

Yes, and this relates to you. You supported the establishment of Columbia’s Women in Computer Science group when you served as Department Chair; this group is now a thriving community of more than 100 undergraduate and graduate students.

You strive for strong representation of women in your PhD group; roughly 45% of your graduated PhDs have been women, more than double the national average.

Can you talk about that journey of forming that group? And maybe there’s some stories that you can share about that process? And where it is today and so on?

Kathleen R. McKeown 35:39

Yes. To think a little bit about what I can say. But yes, when I was chair, it was an undergraduate woman who came to me and said she was interested in forming this group of Women in Computer Science, and would I support it? I said, definitely yes. This was an effort that came about from students’ initiative.

There were several things that surprised me when it started. The first was … and here’s where I think it’s important to have senior women in the room. I’ve seen that, from my own impact. But when I went to the faculty to say that, there were undergraduate women who were interested in forming this group of Women in Computer Science, and I thought there would be nothing but support for that. I found a surprising amount of questions about that; about why was this needed? And would men be excluded? All sorts of things like this. Of course, men in the department, are very supportive of this group. Particularly, as it has grown. It’s been able to accomplish quite a bit. But at that beginning point in time, I saw how important it was to have someone senior who could shield the undergraduates from the kinds of reaction; the initial reaction that came back. And that I could argue for them; and that we could go forward.

That also indicates, sort of the importance of, for bringing younger people in; having more senior women there to support them. That group has been run by students and grown by students. It’s really been remarkable. They’re very passionate about what they do. It provides career advice; panels for what kinds of careers; for how to navigate going through the undergraduate and graduate program. It provides social events for them to get together. We have networking events, each year. It really, I think, has been very important for supporting with the women in the department and the growth of the number of women in the department.

For my own group, I do look for people who are applying and who are coming in and I think about the balance of my group. Each year, I think about the number of women and the number of men I have, and of course, I am always looking for the best to join but if you look you can sometimes, if you look in places where you don’t expect, you can find the best, but it does require paying attention.

Yes, I’ve had some years; I’m coming out of a period of time. I had in the past, like five years and so, I had five women and two men; my crew. Now I’m going back, I have more men right now. I think I have about four men and three women right now. The balance changes and different points in time …, but it is something that I pay attention to.

Stephen Ibaraki 39:50

You have so much experience and wisdom. Collective wisdom from the communities you’ve worked with, and then from the span of your research because it’s so far reaching.

Can you share some insights? What are some of the attributes that you think make for research success? Or maybe that’s just too hard of a question to answer?

Kathleen R. McKeown 40:12

No. There are things I can say about it.

I think a big thing that it takes is determination. The most successful students that I have are those who are really driven. I want to say aggressive in their approach to research, although I know, aggressive is sometimes viewed as a negative word. But they really take it upon themselves to push themselves each and every day. They don’t take no for an answer. They’re willing to put huge amounts of effort in.

Stephen Ibaraki 41:07

You find that there’s sort of a real innate curiosity beyond their field? Or do you find that they very focused in a particular area?

Kathleen R. McKeown 41:17

A little bit depends on the student. That can vary. It can be either way.

I have some students who are interested in the number of different things and will essentially, like move from one project to another. There’s a theme, usually between them. As a PhD student, you really want to have a theme, because you want to have something coherent in the end.

And then in contrast, there are those who are really focused. So, something that’s a matter of personal taste on them.

Stephen Ibaraki 42:05

Do you find your graduate students by and large continue then into research and academia?

Or are a small percentage, maybe 10% or 20%, maybe larger, they say, I want to go into the industry; I want to work for some of the tech companies out there, or other companies.

And some may be entrepreneurial, where they want to form their own companies. Do you have kind of a sense of what the numbers are roughly on that?

Kathleen R. McKeown 42:30

I would say more students go into industry than into academia. I’ve graduated; so, I did this count a couple of years ago. It’s slightly out of date. At that point, I had graduated 37 students, and 14 of them had gone into academia. So slightly less than half.

Stephen Ibaraki 43:01

We covered the span of the different areas that have occupied your interests, research interests, your contributions to enabling more women access and support and so on.

Are there areas other areas that you want to talk about that we missed and so far in our chat?

Kathleen R. McKeown 43:23

In my work, you mean?

Stephen Ibaraki 43:26

It could be any topic now. Your work plus activities outside of your work?

Kathleen R. McKeown 43:35

I do have activities outside of my work. I’m not all about work. So, my biggest passion is sailing. I have just come back from a three-week sailing trip where we sailed from the end of Long Island, to Maine and back. Hit some pretty wild weather along the way. My other passion in life is sailing.

Going back to work, though, I do find one of the things I like about being at a university is that you can shift your focus at different points in time. I could focus more on research. I could focus more on teaching. I could focus on service. At different points in time, I have gone in these different directions.

So for example, I spent about three years where I led the Commission on the Status of Women at Columbia and our focus during those three years was on passing a parental workload relief, before that parental workload relief, there was only the basic maternity leave, which was a six week, disability leave.

That took quite a bit of work and working at all levels of the administration. It had to be passed through the University Senate. But the different Deans of the different schools had to be on board, the president of the university had to be on board.

I learned a lot during that period, like how can you get a university to change? What does it take to get a university to change? And the fact that we did change, there was a lot of resistance in the beginning, and it took a lot of work to get that through. And I’m actually quite proud of that accomplishment. It has impacted a lot of people since it went through.

Stephen Ibaraki 46:09

Well, that’s a significant contribution.

I just have two more buckets of questions.

This next one is about the awards you won. What they mean to you? And then which ones for various reasons, and you could maybe just highlight one or more; just give your viewpoints. I’ll list some of them for the audience.

Kathleen R. McKeown 46:33

Can I interrupt for just a second? Because there’s one other aspect which I should have answered to your previous question.

That is in recent years, I have become an Amazon scholar. I’ve been working at Amazon, which I started during my sabbatical. And I wanted to do it. So I would see the environments that my students would have when they graduated. That has really enabled me to see a different part of the field and how people work and has been quite interesting.

Stephen Ibaraki 47:12

Yes. In fact, that’s really fascinating. Because you’ve had, and will continue to have, many, many key roles at your school and with the community as well. I encourage the audience to look at your extended profile. That’s part of the interview. It’s just fascinating, all the work that you’ve done.

I’m just going to detail some of the awards you’ve won. And then we’ll spend a few minutes talking about that. And then finally, end with some recommendations that you want to give to the audience.

So, first of all, just awards, there’s just so many; you have received numerous honors and recognitions

American Academy of Arts and Science elected member, American Philosophical Society elected member, American Association of Artificial Intelligence Fellow, a Founding Fellow of the Association for Computational Linguistics and an Association for Computing Machinery Fellow. Early on you received the National Science Foundation Presidential Young Investigator Award, and a National Science Foundation Faculty Award for Women. In 2010, she won both the Columbia Great Teacher Award—an honor bestowed by the students—and the Anita Borg Woman of Vision Award for Innovation for her work on summarization of news. And this latest one. I’m interviewing you about the 2023 IEEE Innovation in Societal Infrastructure Award.

There are just so many accolades and recognitions of what you’ve done. You want to make a comment of what that feels like to be awarded or you want to single out any of them. You have an open mic, so to speak, of how you want to talk about the awards you won.

Kathleen R. McKeown 49:00

Well, it certainly feels like an honor. I feel very honored to have received these.

I think the one that probably well; there are two that are probably most meaningful to me and I actually received them both in the same year.

One was the Anita Borg Women of Vision Award. I like being honored by a group like that. But I also liked, sort of the specific award Women of Vision. It’s not something that I would have really been able to say about myself that I was a woman of vision. It’s like, I couldn’t imagine being worthy of something like that. So, it really meant a lot to me. And coming from that community, it meant a lot to me. I very much enjoyed the award ceremony. It was a large audience of young women. And I think they were high school age as well as undergraduates. So you know, you had a lot of women coming up to you afterwards. I always enjoy talking with young women. I love to be able to give advice. That award was very meaningful to me.

Another one, the Columbia’s Great Teacher Award, which comes from the students and again, because that comes from the students. It feels particularly meaningful. I am very passionate about helping the students. I love working with the students. I love working with them at all levels, and, graduate and undergraduate. The students that we have at Columbia are just so amazing. I often feel I learned from them. They are so smart, and really just so special. So to receive an award like that was also very meaningful to me.

Stephen Ibaraki 51:48

You’re talking about it with such passion, that it’s inspiring. I thank you for that; for sharing that narrative.

This is my final question. I’m going to set it up … and that you are the Henry and Gertrude Rothschild Professor of Computer Science at Columbia University and the Founding Director of the Data Science Institute, serving as Director from 2012 to 2017. You mentioned the Amazon Scholar role. You served as Department Chair (1998-2003) and as Vice Dean for Research for the School of Engineering and Applied Science (2010-2012), and of course, your PhD work from the University of Pennsylvania in Computer and Information Sciences in 1982 …

So that gives the added sort of context. And my last question is, based on everything we talked about today, what are your recommendations for the audience?

Kathleen R. McKeown 52:47

A big question. I guess my recommendation for the audience. I would have two kinds of recommendations.

One is for the younger people in the audience, that to pursue your passion, do what you love, and enjoy your work. So, pick something that you enjoy. I have really enjoyed my work over the years.

And then I guess, for the broader audience is just, I think, now is the time when natural language processing has had a huge impact and has the possibility for far more impact going forward. I enjoy seeing it. I encourage efforts in all areas where it can touch whether it’s more on the business side, but also on the societal side where we can make a strong impact to help humanity.

Stephen Ibaraki 54:00

Well, I can see definitely the work that you do for the benefit of Earth’s ecosystems which includes humanity.

Thank you so much, Kathy, for coming in today. You just have an outstanding, amazing, career … contributions that are felt worldwide and will continue for some time.

Thank you for sharing your insights, wisdom, your experiences, your narratives and stories with our audience.

Kathleen R. McKeown 54:25

Okay, well, thank you so much for having me.

Advertisement

Leave a Reply

Your email address will not be published.