Phase III

[ Overview - Phase I - Phase II - Phase III ]

For Phase III, you will need to propose an extension of your project or application using it. Examples of extensions include advanced models, faster search algorithms, advanced compression, handling dynamic collections, hyperlink algorithms, etc. An application would include a more usable front-end for your search engine that allows the user to employ the engine for some task, such as collecting bookmarks, automatically indexing email, etc.

You will write your proposal (length should be a page or two) and discuss it with the instructor for approval. Your proposal should include (a) what you plan to do, (b) what milestones you will traverse to accomplish it, and (c) how you will evaluate or assess the success of your extension, in the form of two or more benchmark tasks in the spirit of Phases I and II. You might not actually do (c), but you should think about how it would be done.

Your proposal should include at least three references on the subject of your proposal. This helps me know that you have read about your problem and have a good idea of how it can be done.

Phase III proposals are due on October 31.

Some possible Phase III projects

(some are more realistic than others... ;-)

Language modeling
Latent semantic indexing
Statistical or linguistic phrase identification, named entity extraction
phrase, proximity, NOT , and "more like this" queries
searching with errors
passage retrieval
Link analysis methods (umbc-crawl only!)
Anchor text (umbc-crawl only)
Relevance feedback, with term selection and an interface
Filtering
Focused crawling
Distributed Retrieval
More coming soon...

Deliverables

In addition to the relevant sections of your design document, journal, code, and benchmark output, you (or your group) will give a 15-20 minute talk in class describing your project, it's (unique) goals, and your results.

If your Phase III output requires a demo, schedule it with me as early as possible.

I expect that many Phase IIIs will not be fully completed by the end of the semester! Many of your ideas will likely require more time than we have left to reach full fruition. I will grade your Phase III on the basis of

Quality of the proposal
Evidence of background research
A best-effort attempt to accomplish the task
An understanding of what it will take to get to the "finish line".

With this in mind, I do not approve proposals which are clearly beyond the scope of the class, or which are trivial implementations of the lecture material. An ideal Phase III proposal takes an idea (from the outside reading or some other source) which can be accomplished in the remaining class time, or slightly longer, by a diligent group of students. Think of it as a proof-of-concept design that could be the basis of a Masters thesis or project.