• a project design including
  • diagrams representing and explaining the data structures to be constructed by the program during the course of the computation
  • charts showing and explaining the transformations the data experiences as it moves from one data structure to the next during the computation
  • a Program Organization Chart exhibiting the interactions between major program components. Attach a report to the chart providing a rationale for the program organization. (Why does it organize the parts of the computation in a coherent way? How does it facilitate independent work on different parts of the program by different people?)
• a listing of the Haskell script the team writes for the project
• a session transcript in which you execute the command "main". This causes the Hugs system to produce the value of the variable "main", defined in the Main module that you write for this project. An effect of producing the value of main is an interactive session with the person at the keyboard.
• printed copies of the files ques.dat and resp.dat
• annotated agendas of team meetings

• Each team must schedule a meeting with the instructor to review the design for their project. The meeting must take place before March 28.

• Interactive Behavior of Program
  The person at the keyboard is prompted to enter answers to questions from a questionnaire. After the person's answers are retrieved and appropriately converted, they are matched against a database of pre-recorded questionnaire responses. Finally, the coded identifiers of the three people from the database whose answers most closely correspond to those of the person at the keyboard, along with the degrees of correspondence, are revealed on the screen.

  The questionnaire is recorded in the file ques.dat. The database of pre-recorded responses (to which the person at the keyboard's response is compared) resides in the file resp.dat.

  The prompts in the interaction are questions from the questionnaire file (starting from the question in the first questionnaire item in the file, continuing with the question in the next item, and so on). Questions should be re-formatted to fit on a 60-space line, left justified, but not necessarily right justified. If a question contains a word with more than 60 characters, that word can appear on a line by itself which, of course, will exceed the limit for line width.

  Each answer must be one of the valid answers designated in the corresponding questionnaire item (except that the comparison should ignore letter-case if either the answer entered from the keyboard or the valid answers from the questionnaire item contain letters -- for example, a keyboard entry of "C1b" should match a valid answer represented as "c1B", "c1b", or "C1B", as well as "C1b"). If an entered answer matches none of the valid answers, then the person at the keyboard should be asked to re-enter the answer.

• Format of the Questionnaire File
  The file ques.dat is a sequence of questionnaire items.
  A questionnaire item is sequence of lines, terminated by an empty line, containing three sections:
  • question - The first few lines of the item contain a question.
  • valid answers - The line that is two lines prior to the empty line that terminates the item contains a sequence of words (separated by spaces) representing the set of valid answers to the question.
  • numeric equivalents - The line following the valid answers contains contains a sequence of words (separated by spaces) representing the numeric equivalents of the valid answers to the question. The number of words on the numeric-equivalents line must be the same as the number of words on the valid-answers line next to it.
  The term "word" is used in this project in the same sense that it was used in Team Project B.

• Format of Response Database
  The file resp.dat contains pre-recorded responses to the questionnaire. It's data is arranged in the same format as the data files for Individual Project 4. The parameter headings will be ID, Q1, Q2, Q3, ... Each data identifier will be the coded identifier of the respondent whose response to the questionnaire is recorded on the line of the file containing the data identifier. (The database uses coded identifiers instead of names to avoid revealing the names of the respondents. Note that data identifiers, as described in the data format of Project 4, contain no spaces. Since the data identifiers described in Project 4 are the coded identifiers of this project, coded identifiers also contain no spaces.)

• Converting an Answer to a Number
  Suppose the valid-answers line and the numeric-equivalents line of an item in the questionnaire file look like this:

       a     b     c     d     e
       0     1     2     3     4
  

  Then the answer "a" would be converted to the number 0, "b" to 1, and so on. In general, an answer matching the n-th word on the valid-answers line is converted to the number denoted by the n-th word on the numeric-equivalents line.

• Evaluating Degree of Correspondence Between Responses
  The degree of correspondence between two responses is the correlation (as defined in Individual Project 5) between the normalized sequences represented by the two responses. The larger the correlation, the closer the correspondence between the responses.

  Normalized Response Sequence

  The sequence of answers in a response is normalized by dividing each of the of numeric equivalents of the answers by the appropriate normalizing factor. The normalized response sequence is the resulting sequence of quotients.

  Normalizing Factor

  The normalizing factor for the answer to a particular question Q is the standard deviation of the numeric equivalents of the answers to question Q recorded in the response database.

Ground Rules

• The program should be able to process a questionnaire with any number of questions, each of which can have any number of valid responses. Nor should the program depend on the number of responses in the database. The questionnaire file that your team constructs for the project should contain at least five questions, and the response database should contain at least ten responses.
• In each team project, certain team members will have special assignments:
  • leader
  • recorder
  • librarian
  • collector
  The Team Organization document outlines the duties associated with these assignments. The people who play these roles will be chosen by the team at the first work session for the project. Team members should play different roles on different projects. Contact the instructor if your team has difficulty rotating assignments.
• You may use any aspect of Haskell to complete this project. Modules in Supplied Software and solutions to previous projects are good sources of useful functions for this project.