Researcher of the Month
April 2009
Boubacar Diallo
Computer Science & Business major, Exchange student
Mentors: Dr. Robert Kelly and Dr. Tony Scarlatos, Computer Sciences
Parlez-vous C, C++, Java, JavaScript, XML, XHTML, CSS, PHPS, RoR?
Meet Boubacar Diallo, an exchange student from The Engineering Institute for Information
Technology and Business Applications who is fluent in French, English, Spanish, and
a host of computer languages; came to Stony Brook from Paris, France to immerse himself
in research; and has done just that!
Boubacar Diallo currently works with Profs. Rob Kelly and Tony Scarlatos in the CEAS Department
of Computer Science, and has made the most of the research-based class work he is
doing this spring semester both individually and as a team member of his research
projects.
Be sure to stop by and talk to Boubacar Diallo about the projects that he is presenting
at the upcoming URECA Celebration on April 29th!
These include:
- Project 1: Text normalization of chief complaint data. A clean Chief Complaint (CC) record is widely recognized as an effective source of
patient-focused clinical data. These data, which are typically gathered from patients
during Emergency Department (ED) triage, are potentially useful in a real-time clinical
trial candidate selection process, but only if this free-text information is standardized
so that it can be correlated with clinical trial candidate selection criteria. This
standardization requires managing word variations that result from use of synonyms,
abbreviations, acronyms, truncations, concatenations, misspellings, and typographic
errors. We review different approaches for managing word variation and discuss their
limitations. We outline a new approach to text normalization on which three approaches
to handling linguistic variation are based. In addition, we examine the extent of
word variation in the context of ED CC data in Stony Brook Hospital, and measure the
suitability of this data for the automatic identification of clinical trial candidates. Research supervisor: Prof. Rob Kelly.
- Project 2: Teamagine. Teamagine combines speech recognition, handwriting recognition, a semantic knowledge network, and multi-touch capability to provide a public forum for brainstorming, and to support group decision-making. This provides a many-to-many relationship where all the participants are involved. Using infrared light pens for input and a projection device to provide a shared whiteboard, Teamagine creates a fluid environment for group brainstorming activity. The solution we provide turns any tabletop or blank wall space into a screen for generating, organizing, and exploring ideas. The concepts discussed in a meeting around the table are visualized and can be organized and manipulated in spontaneous natural ways by several participants. Users control the application and add objects to the idea map by speaking, add labels to the map with hand written notes, explore related concepts by tapping them, and group related ideas by dragging their graphical representations. Users can even draw thumbnails of their ideas. Queries can be passed to commercial search engines. A database back-end allows users to save and bookmark brainstorming sessions. Research supervisor: Prof. Tony Scarlatos.Group/team members: Boubacar Diallo, Gary Wong, Steven Yu , Jory Osterman, Omari Joseph.
Born in Fria, Guinea, Boubacar moved to France at the age of 7. He completed his Baccalauréat at La Mennais-Saint Armel in Ploërmel, France. While pursuing a software engineering degree at The Engineering Institute for Information Technology and Business Applications, Boubacar also spent a few months in 2006 at University College of London, U.K. studying mathematics and networks, and has had several summer experiences working in IT and software development. Boubacar is currently a phase 3 finalist in the DARE competition for new entrepreneurs: if funded, he hopes to launch a new web-based business for recruitment agencies together with some colleagues here and in France. A fan of Paris, London and New York, Boubacar also hopes at some point to return to the States to work in the New York City area. Below are some excerpts of his interview with URECA Director, Karen Kernan.
Karen: How did you get involved in research at Stony Brook ?
Boubacar: As an exchange student, I knew about Stony Brook when I was in college in Paris.
Since the first year there, I asked about the exchange program. I wanted to be in
New York and I knew that Stony Brook was good, perfect for what I wanted to do with
research. So I applied for the school... I’ve worked in projects in companies or in
school, but I never did a research program before. In the projects I worked on before,
the framework was always already there. We had the instructions. I wanted to see what
it feels like to do a research program when you don’t have all the guidelines, when
you don’t know exactly where you’re going. I wanted to try that. I had classes with
both Prof. Scarlatos and Prof. Kelly as teachers last semester. That’s how I knew
them, that’s how I had a good feeling about them and found out about the research
classes.
You were one of a handful of students who actually submitted TWO projects for the
upcoming URECA Celebration. Both sound great! Tell me a little more about them.
Prof. Kelly gave me several subjects for research. I chose one of them, related to
a specific part of a big project concerning chief complaint data taken from patients in an emergency room. It’s a little piece of data, usually
3 or 4 words, max 5 words, taken by the nurse or physician when the patient comes
to the ER. For instance, if a patient came in with abdominal pain, in some hospital,
a nurse or physician would write an abbreviation for “abd.” It has been recognized,
that if we could use this data, we would have the history of the patient, the problem
he/she had. We would be able to use statistics on how many patients have this problem,
what is the frequency of the problem, etc. But the problem with this little piece
of data is that it is not standardized. There is no conventional way to take an input
of this data. We don’t have a way to uniquely identify a word. So my job is to come
up with a method, an algorithm, to clean this data.
And the second project . . ?
Teamagine is a project developed in a class with Prof. Tony Scarlatos. I took Introduction
to Multimedia Systems (CSE 334) last semester. I really liked this class. He told
us at the end that he had another class, called Human Computer Interaction (CSE 323)
where we would apply the principles that we learned. I was really interested and excited
about this project so I took the class. We had to come up with several project proposals.
I came up with this new project — a tabletop or shared whiteboard that will allow
multiple users to brainstorm about a project. You might have seen the Microsoft Surface
multi-touch table, or the movie Minority Report? I wanted for our project, this kind of table where all the collaborators are around
it and interact with each other. Each collaborator armed with an infrared pen, see
the other people’s ideas and can interact with them. I’m a really visual person. Indeed
I like to have diagrams in class material, and write down notes about what I’m reviewing
for exams etc. Therefore I thought having a visual representation of what you are
saying or thinking about during a brainstorming phase, would help to have more ideas.
The goal is, instead of a one-to-one relationship (user and computer), to have many-to-many
relationships (users around the table).
Is this going to be difficult to implement?
I hope that we will have the demo working for URECA Celebration. But we don’t know
exactly. Spring break makes us lose one week. It’s a lot of time. We will see...
I know you will be presenting the research work at the URECA Celebration on the 29th.
But up to now, how much experience have you had in public presentations? Do you have
regular group meetings with the two projects in which you're involved?
For the project with Dr. Kelly, every Monday morning we have a group meeting during
which we say what we did during the week and what is our objectives for the next meeting.
If there is an urgent problem, we talk and get inputs in order to solve it. What I
really like, especially when we are doing those projects, is the teamwork and the
interaction with other people. How you learn from other people. You always have something
to learn from other people: some are good at programming; some are good in working
methods and other good in communication etc. I always try as much as possible to be
involved in efficient teamwork.
Teamwork seems especially important in computer science.
Definitely! There is no project that you can do by yourself. Indeed it’s time consuming
and error prone. You must have a sense of teamwork. For the other project with Prof.
Scarlatos, I’m the project leader. We have classes on Monday and Wednesday. So right
after class, we have a meeting of at least 30 minutes; and every other day we meet
again to see where everybody is on the project and what our next objectives are. We
try to get ahead, and to always think about the project as overall: is it good in
term of design? Can we make enhancements? How can we do it? We have a lot of interaction
between us and with the teacher. Everybody is really motivated. We are all working,
and it’s a really good team.
You sound very positive!
Yes - because for both projects, we have good teamwork and good relationships with
our mentors. I think if you are doing a research program. . . that means you have
a little bit of interest. You are curious, autonomous — willing to work. Otherwise,
you wouldn’t apply for the program or wouldn’t be in the class.
How beneficial has it been for you to do research?
I knew before coming that Stony Brook is ranked as one of the best colleges when it
comes to research. What I’ve learned from studying here and doing research, is that
the process of doing research is really eased. If you have an idea, what is really
good here is that you have all the equipment and facilities to do it. The research
is really emphasized in a way to help you build and create your project. You have
the faculty members available, a department that help finance the project (for instance
the Wiimote and infrared pens for Teamagine) so you can actually implement your idea.
You have all these people supporting you.
What have you learned specifically from your mentors?
Let me start with Prof. Scarlatos. His class last semester really gave us all this
sense of creativity, the sense of being curious, being on the edge of technology,
the latest ways of interacting with computers. He always pushed us to think: what
is a good design? Do you see it as fun for the user? Do you think they will like it?
He was pushing us to come up with the best ideas. So I learned that from him — being
able to push the limit and try to come up with something new that is fun for the user
and is also fun for people creating it. . . . From Prof. Kelly, what I learned is
how to build and provide good presentation. When he was doing his classes, I really
liked them. He was straightforward to understand. You had everything clearly laid
out in the beginning, your learning objectives, class content, diagrams etc. When
he was in class, or in our research meetings, he was always asking us to ask ourselves
the right questions. This is what I learned from him: being structured in the presentation
that you do, and to ask yourself the right questions.
You mentioned before that you had been in involved with work projects and projects
for classes …but that doing research was a different kind of experience.
Exactly, because when you are in an academic program, or in a company, you have some
big guidelines and you know what you are supposed to do. I would say in a research
program, you are not left by yourself entirely. You do have a mentor who helps you.
But he’s more flexible, and lets you come up with hypothesis, problems, questions
and you find together the best solutions. What is interesting is to be able to be
autonomous and to know that the research will go as far as you involve yourself in
the research. You are directing the research. That means, the more you do, the more
you will get as a consequence of that. What is a little frustrating, but not too much,
is that whenever you are doing research, you never know if it’s good—if what you’re
working on is the right thing. You ask yourself: how long should I invest time in
the researching phase? When should I come up with something that works, that can be
implemented?
Have you always been interested in computer science?
Yes – but I can’t really explain it because I feel like it’s always been there. I
knew since I was 10 that I wanted to do something in computer science. Whenever we
had any new electronic equipment — DVDs, TV, phone — I would be the one to help set
it up. I would read the manual even if it was long sometimes. When we had our first
computer, I helped to set up everything.
It sounds as if you really enjoy the process of setting up and doing problem solving
. . .
Yes, it’s part of the job. It shows your ability to manage things in critical situations
that you will encounter in your future work. If there is a crash in the system or
a maintenance problem, you need to quickly know how to fix it the most efficient way.
I see from your resume that you've travelled a lot, and know several languages. And
it looks like you’ve also had to learn lot and lots of computer languages!
When you’re doing programming, the most difficult is to create algorithm, the method,
the process, the concept. When you learn a new language, it’s like learning a new
language like French, Spanish, and English. You just learn the syntax, how to write
this idea in this language. It’s pretty straightforward. And since it’s something
I really like and enjoy doing I can spend a lot of time on it. Just recently, I learned Ruby on Rails by myself. I’m just curious so I will try to learn what I can, what will be useful.
I think that what I always try to look for, especially when training to be an engineer,
is to be able to adapt to any situation. Especially in computer science field, there
is a new technology very often and as future software engineer, it’s our duty to be
aware of them. So I don’t try to be an expert in a specific field, but rather to have
knowledge in many different fields. What they teach us in school, both in France and
here, is how to teach ourselves. Especially in the computer science field, we need
to combine our knowledge to another field: medical field, physics, biology, manufacturing
etc. When you need to learn something specific about an unknown field, you just teach
yourself in a short amount of time to be able to work on the project.
With your chief complaint data project supervised by Prof. Kelly, you must have had
to learn a lot about the emergency room details of operation....
Exactly and I like this interaction with people. For this project, I interviewed nurses
and physicians by asking questions about on their job on a daily basis.
Communication seems to be one of your strengths!
I don’t know if it’s a strength, but it’s something that I like doing. I think it’s
really important. You are not alone in any job. You really need to be able to communicate.
In a business, or in school, you are working with people with different background,
and you need to have a broad view of what everybody is doing by asking the right questions
to the right persons at the right time; and at the end of the day, come up with solutions,
something that’s going to work.