Project 1 Proposals

by Mike Gleicher on February 29, 2012

What a project proposal must have (examples/details follow):

  • A Descriptive Title (a sentence phrase that gives an idea of what the project is about and what you are going to do)
  • A Brief Description (a few sentences that explains what you are going to do)
  • Pedagogical goals (a list of things you hope to learn by doing it)
  • Milestones (what do you expect to have at the steps along the way)
  • Expected outcomes (a list of the things that you hope do / deliver)
  • Resources (a list of the things you will need/use. this includes the starting points for literature explorations)
  • Evaluation Plan (a few sentences on some of the ways you will be able to know if you have succeeded)

Remember, this is a plan, not a contract. However, it is useful for us to “negotiate expectations” now, and for you to have a good sense of direction. Adjustments can be made as you do the work and better understand where you expect to end up.

As I count, this is (approximately) a sentence + 2 paragraphs + 4 lists of 2-4 sentence/phrases + another few sentences. Maybe a page or two.

I really believe that thinking through your project well enough to write this will raise the probability of a good outcome.

Please send this via email to the TA and the Professor. You are not “approved” until one of us sends you an email that explicitly says that you are approved. (that email may come from either of us). We will make best efforts to give you prompt replies.

Descriptive Titles

Here are some examples:

A Literature Survey of Approaches to Scalable Data Display Organized by the Gleicher’s Strategy Taxonomy

Notice how this implies what i am going to do, as well as the topic. “Gleicher’s Strategy Taxonomy” is something from a paper

A Design Study of presenting motion trace data from manipulation experiments

Again, it’s clear what kind of project this will be and what the topic area is

Comparing toolkits for the web-based implementation of scatterplots

Here, it’s a little less clear – but there’s an implication that this is going to mix some survey review, some design study, and some implementation.

A Brief Description

In a few sentences, describe what the project is about and what you are going to do.

The topic should give a sense of what the problem you are facing is. Be concise (a few sentences), but it it probably worthwhile to say something to frame the problem and explain why its interesting

Scatterplots are a very common chart type, and have a diverse range of extensions including interaction technqiues for exploration, visual techniques to show additional data dimensions with the points,  and aggregation techniques to deal with large data sizes. These extensions mean that flexibility is important in a scatterplot tool, as most uses will require some extensions. In this project, I will examine several web-based scatterplot tools to see how flexibly they can accomodate extentions.

The second part should be specific and involve the verbs and tangible things you will do.

In this project I will survey the literature and look at examples of interactive scatterplots to catalog the common kinds of needs that scatterplots need. I will examine the documentation for 3 common web-based toolkits that allow for scatterplot construction (D3, Google Charts and Protovis) to understand how easily they accomodate extensions. I will implement small examples of extended scatterplots (such as using different shape encodings) in these three different toolkits. I will compare how easy it is to create basic scatter plots as well as to extend to richer scatter plot features.

Pedagogical Goals

This should be a list of the kinds of things you hope to learn by doing the project.

Some examples off the top of my head…

  • I want to get familiar with the range of techniques that address a particular problem.
  • I want to learn to organize a survey of methods into a useful taxonomy.
  • I want practice at a problem-focused design methodology.
  • I want to learn to program in Javascript.
  • I want to learn how to implement systems that scrape large amounts of dirty, poorly organized data.
  • I practice at designing displays that pack a lot of information into a small space.

Milestones

For each week, give a few concrete “check points” that you can see if you are on track. One way to think about it: if you don’t make this checkpoint, it’s a sign the project is behind and you may have difficulty getting to the endpoint.

Be sure to leave time for writing at the end!

Some examples (note: these are all from different “projects”)

  • Week 2 (March 9) – I will have read 2 survey papers and have collected a list of all the other papers that I want to read.
  • Week 3 (March 16) – I will have implemented some basic things with the tools, and have some program that reads in the data files.
  • Week 4 (March 23) – I will have sketched out and documented at least 5 designs to evaluate against the tasks I had identified in the earlier phase.

Expected Outcomes

What do you expect to produce at the end. Note: that ALL projects will require some written document, a list of references used, and a (potentially empty) list of other references that you identified but didn’t read.

Especially if you are doing a survey or design study, you will probably accumulate larger number of papers and/or examples to look at than you will examine in detail. So you might produce 3 lists – things I looked at carefully, things I looked at quickly (read the abstract), things I identified but didn’t get to.

Describe each of the things you expect to have when are finished.

Some examples of descriptions of project components:

  • I will have a document of 2-4 pages that defines the task problem and task needs.
  • I will have a gallery of 3-5 designs, each shown with a few Illustrator drawings, and a page of text describing their operation and explaining their rationale.
  • I will have an annotated reading list, with brief summaries of the primary references, and lists of other papers that I found that was not able to include. (note: a survey is more than just an annotated bibliography it needs to to organize and synthesize from this set of readings – it may include an annotated bibliography).
  • I will have a prototype interactive system that works on a sample data set. (note: this is explicitly not claiming to have a general system that works on multiple data sets).
  • I will have an off-line prototype that will produce an image from the sample data set, as well as a mockup of what the user experience might be. (note: this is assuming that generating the visual design is sufficiently hard that deferring a nice implementation is sufficient).
  • I will create a 3-5 minute video that shows how a mockup of the design can work using various sketching/prototyping tools.
  • I will create a set of guidelines that describe how to choose among the various designs that I document.

To emphasize: there is no set expectation of length. The diverse range of students and projects means that what gets produced will be different in each case. Also, expectations are lower for 638 students.

Also, your project will probably involve a mix of outcomes. All projects must include a reading list, and some write up.

Resources

Describe the various things you will need in terms of the project.

  • If your project involves working with a data set, you must explain what the data set is and confirm that you can actually gain access to it.
  • If your plan involves implementation, describe the tools you plan to use, and confirm that you have access to them. If you want to build something with a server back-end, you may need to provide your own.
  • You should list any initial “footholds” into the literature that you have. If a literature review is an important part of your project, you should discuss how you will find “enough” (where you get to define enough).
  • If your project involves examining existing designs, you should discuss the designs that you plan to look at. While you may not have the complete list, you should at least have some, and you should have a plan for how you are going to find “enough.” (where you get to define enough).

Some examples:

  • I will assemble a list of literature by reading over the abstracts for the past few InfoVis conferences, the past few years of the Information Visualization journal, and the references from the survey papers by Elmquist et. al and Ellis and Dix.
  • I will implement the prototypes in Javascript using JQuery and D3. The data will be put in static files that can be served locally or from a website without server-side programming.
  • I have found these 3 examples of interactive web-based visualizations on the NY Times web site (with links), and will find more examples around the web by reading through blogs including (list of things).
  • I have found a dataset of things on website X and have downloaded it. There is another dataset on website Y, and figuring out how to scrape it will be one of the key implementation components of the project.
  • Professor X in the Chemistry Department has given me data for 10 samples, and has promised to give me 50 more next week.

Evaluation Plan

Write a few sentences (maybe as a list) giving some ideas about how you will know you are successful. What properties would you look for in the outcome that would make you say “yes, this really deserves an A.” What would a successful outcome look like?

Anything Else

If you want to submit more, feel free. Things like initial prototype designs, initial reading lists, examples of existing visualizations for the problem you’re interested in, …

Please attach this at the END of the document. (or as a separate document). While we will try to give you feedback, it might come after we’ve gotten through giving everyone an OK.

{ 1 trackback }

Previous post:

Next post: