Kimberle Koile, Ph.D.

Research Scientist
MIT Center for Educational Computing Initiatives

9-323
(617) 253-6037
kkoile at mit.edu


 Building  location:  9
                                                                                                         

Projects

Educational Technology:  Classroom Learning Partner

Streetscape:  Qualities of Urban Context  

TAC:  The Architect's Collaborator

835:  Intelligent Workspace


MIT + RSC:  Technology and Theatre


Teaching

6.001:  Structure and Interpretation of Computer Programs

6.034:  Introduction to Artificial Intelligence

6.871:  Knowledge-Based Application Systems

            Previous Industrial Work

            Arris Pharmaceutical


            Unisys

            BBN



Projects

Educational Technology:  Classroom Learning Partner
     
 
The goal of the Classroom Learning Partner (CLP) project is to improve student learning in large classrooms by increasing instructor-student interaction and student learning.  In particular, we'd like to turn lectures into more of a two-way conversation between instructor and student than has been possible to date.   The two-year project, which began in June, 2005, employs Tablet PC technology to support hands-on activities in the form of in-class exercise:  The CLP group is extending an existing Tablet-PC-based presentation system, Classroom Presenter, to support the submission and aggregation of student solutions to in-class exercises (Anderson, et al. 2004).  Using the current Classroom Presenter system, an instructor lectures using a wireless presentation system, annotating slides with digital ink.  The slides and ink are displayed simultaneously on a large screen and on students’ Tablet PCs.  An instructor displays a slide containing an exercise for the students to work.  The students work the problem, using digital ink on their tablets, then anonymously submit their answers to the instructor via a wireless network.   The answers are collected into a new slide deck on the instructors machine, and the instructor can select individual answers for class discussion.  Using Classroom Presenter in this way works well in classes of size eight or smaller (Anderson 2005), as instructors are easily overwhelmed by more than eight solutions. 
     CLP extends Classroom Presenter so that in-class exercises can be used as a teaching method in large classes:  Using AI methods, it interprets and aggregates student solutions into a small number of equivalence classes and presents the results to the instructor, e.g. in the form of a histogram and representative solutions (Koile, et. al 2007).  
     
   Anderson, R, Anderson, R, Simon, B., Wolfman, S., VanDeGrift, T. and Yasuhara, K.: 2004, Experiences with a Tablet PC Based Lecture Presentation System in Computer Science Courses, in  SIGCSE 2004.
  Anderson, R: 2005, personal communication with K. Koile.
  Koile, K., Chevalier, K., Rbeiz, M., Rogal, A., Singer, D., Sorensen, J., Smith, A., Tay, K.S. and Wu., K.: 2007, Supporting Feedback and Assessment of Digital Ink Answers to In-Class Exercises.
In Nineteenth Conference on Innovative Applications of AI, July 22-29.

  CLP Publications:

     Please see the CLP website.


Streetscape:  Qualities of Urban Context
    This study investigates a computational method  for representing design knowledge of urban context and using that knowledge to evaluate qualitative characteristics of urban streetscapes.  The study focuses on streetscape as an urban element, in particular on the relationship between physical characteristics and experiential qualities of  the pedestrian environment.  The design exercise is to understand and manipulate the physical characteristics of the street space so as to realize specified experiential qualities, e.g., approachable.  The project goal is to build a prototype system that evaluates experiential qualities of a given urban plan and recommends modifications for better achieving specified qualities.   The system will be used to investigate an approachable fence system for Yonsei University, Seoul, Korea, and a revitalized neighborhood for Harvard Square, Cambridge, MA.  (Student advisee:  Jie-Eun Hwang, Harvard Graduate School of  Design)

    Streetscape Publications
:


TAC:  The Architect's Collaborator
    
The Architect's Collaborator (TAC) was built to explore two questions: How are abstract terms such as “private” and “open” translated into physical form? How might one build a tool to assist designers with this process? TAC is a design assistant that supports iterative design refinement, and that represents and reasons about how experiential qualities are manifested in physical form. It explores the space of possible designs in search of solutions satisfying specified goals by employing a strategy we’ve called dependency-directed redesign: it evaluates a design with respect to a set of goals, then uses an explanation of the evaluation to guide proposal and refinement of repair suggestions; it then carries out the repair suggestions to create new designs.

    TAC Publications:  
835:  Intelligent Workspace  (not active)
     
 
To build workspaces the effectively support people in everyday work situations, one needs to focus on both the technology and the physical environment in which the work takes place.  In building our intelligent workspace, the aire group (agent-based intelligent reactive environments) is focusing on three areas of research. First, we are exploring the use of mobile furniture to create a dynamically reconfigurable workspace:  as the physical arrangement of the space changes, the technology supporting the space should accommodate accordingly.  Secondly, we are focusing on tools to enable this sort of dynamic configuration:  we are developing novel computer vision and other sensing technologies, in collaboration with the Vision Interface Group , for inferring activities in a space.  With knowledge of the activities in which a person is engaged, an intelligent workspace can offer relevant assistance.  Our third area of research focuses on the study of how people use an intelligent workspace such as the one we are building.  (Student advisee:  Matthew Morrissette)

  835-related Publications:
 

MIT + RSC:  Technology and Theatre (not active)
   
MIT and the Royal Shakespeare Company are in the beginning stages of a collaboration, the aim of which is to explore the uses of technology in new approaches to rehearsing, performing, teaching, studying, and experiencing Shakespeare's work and other classical theatre. The campus-wide collaboration focuses on four areas:  rehearsal and performance, educational media, live experience through games, new works. Our group is focusing on rehearsal and performance and will explore the use of technology to enhance the creative process in rehearsal and the experience of performance, both for actors and audiences.  Members of the MIT Comparative Media Studies and Theatre Arts Departments are focusing on:  educational media,  creating an interactive documentary to explore new RSC productions from concept through early development, resident production, and tour;  games,  developing video games in which players are introduced to Shakespearean worlds and narratives by being immersed in a simulated world where they create and improvise with characters in Shakespearean plays; and new works,  developing a production that explores the creative process and examines the differences between creating new works in the 17th century and now.  (Student advisee:  Andrew Perelson)

    Business week article, May 26, 2003:   "Lights, Hologram, Action."
    Andrew Perelson's MEng thesis, May, 2005:  "iPlot:  An Intelligent Lighting Design Assistant."




Teaching

6.001:  Structure and Interpretation of Computer Programs  (Spring 2005 to 2007

6.034:  Introduction to Artificial Intelligence
  Co-taught with Prof. Patrick Winston.  (Fall 2002, 2003; Fall, 2007)

6.871:  Knowledge-Based Application Systems
  Co-taught with Dr. Howard Shrobe.  (Spring 2003, 2004)


Previous Industrial Work

Arris Pharmaceutical:  Senior Computational Scientist, Computer-Aided Drug Design Group
    Leader of a 10-person project that developed and integrated data analysis tools with data-generating laboratory equipment for synthesis and assay of compounds.  Liaison between computer scientists and chemists.  Designed overall system architecture; developed domain models for the drug design process, as well as representations for chemical and biological knowledge; developed a user language based on the vocabulary of the drug designer.  Member of project team that developed machine learning tools for predicting biological activities from molecular surface characteristics.  Responsible for representation of molecular and biological knowledge.


    Arris Publications:  

Unisys:  Research Scientist, Knowledge-Based Systems Group
    
Designed and implemented a system for reasoning qualitatively about gene regulation.  Built a knowledge base of molecular biology, along with associated knowledge representation tools.
 
    Unisys Publications:  
BBN:  Research Scientist, Intelligent Systems Group, Life Sciences Group
    
Designed and implemented knowledge base and user interface for an intelligent system for communication network performance analysis.   Designed and implemented a system for protein analysis and design that enabled protein chemists to analyze and compare secondary and tertiary structure of proteins.  Designed and implemented a graphical interface for a system that aided chemical engineers in pharmaceutical plant design, and an interface to a natural language system (IRUS).  

    BBN Publications: