Adwait Ratnaparkhis Dissertation Proposal

(Adapted [ ], with permission and grateful thanks, from Leonard Cassuto, Fordham University.)

The purpose of a research proposal is for it to be approved. Only then can you start writing. A lot of misunderstanding swirls around research proposals. One foundational fact cuts through it: A research proposal has no independent existence. It's a provisional document, a way station to an eventual goal.

A research proposal is not an essay. In the humanities and some of the social sciences, a proposal looks a lot like an essay, but it differs in one fundamental respect: While an essay must prove a thesis, a proposal needs only to advance one. It's enough, in other words, for a proposal writer to demonstrate an argument and show how to prove it at a later date-given approval, space, and time.

A research proposal is not a mini-dissertation. If a dissertation is a small world that you (as god of the microcosm) will bring into being, a proposal is a map of that space within the larger universe. The emphasis here is on the idea of mapping rather than creating. Before you can become a god and invent your own world, you have to become a cartographer.

That means that the goal in your proposal is not to create your world, but rather to suggest what it will look like when you do create it. Because you're mapping a world that doesn't exist (and here my metaphor becomes strained), you should imagine that you're diagramming a place you haven't been to yet.

It's a common mistake for a proposal writer to fall into writing the actual dissertation in the process of laying it out. That's not entirely a bad thing: It offers you a head start. But because students too often misunderstand the nature of the project, most research proposals take too long to complete. The proposal is a provisional document that marks a point of transition, not a polished work of compressed scholarship that need only be inflated to become a dissertation.

A proposal describes your project from both inside and outside. First, the inside stuff:

A proposal puts forth your argument. It points toward how it will be proved, perhaps giving a few well-chosen examples but without unspooling them in detail. A few exemplary details will help illustrate your presentation, but a profusion of them will distract. Such details serve the purpose of demonstrating-not fully proving-your argument.

A proposal describes how your argument will fit together. What examples will you use, in what order, and why? How is the argument sequenced and subordinated?

A proposal outlines methodology. How will you make your argument? What theoretical, historical, contextual, and interpretative tools will you use? Will you employ any particular approach?

Your proposal should fit your dissertation topic. A proposal to edit a scholarly edition, to pick one exceptional possibility, will require a different presentation than a dissertation laid out in the model of a monograph (introduction plus four chapters on related topics). The shoe must fit the foot and not the other way around.

From the outside:

You need to show the place of your dissertation in the critical field. Which field and subfield conversations will your project enter, and how? Which critics will you be building on, and which ones will you be revising? Your dissertation marks your formal entry into the community of scholars, a world of intellectuals engaging in overlapping conversations of varying size and scope. Your proposal must show your awareness of those multiple discourses and show the place your research will occupy within them.

Accordingly, you should include a thorough bibliography in your proposal so that readers may look at what works you plan to consult, as well as those you have consulted already. Staff reviewing your application will review that list and use it as the basis for further suggestions.

Finally, I offer proposal writers a commandment and a postulate.

The commandment: Consult prospective supervisors as you develop your proposal. The myth of the writer as solitary genius striving away in the garret has surprising persistence. I've seen many graduate students teach their undergraduates to collaborate without realizing that they're not following their own advice. (That is a mistake I made often enough myself.) You should not imagine that you will be writing your proposal on your own. Instead, draw on the experience of your peers, and especially  of prospective supervisors, as you shape your topic so that it may be the most relevant, the most challenging, and the most marketable later on.

The postulate: Your dissertation will be different from your proposal. That's to be expected-and the differences can be substantial. Your proposal outlines a hypothetical dissertation: what your thesis looks like to you from where you stand now. The goal of a proposal is not that it should outline your future dissertation. Rather, it should outline one possible dissertation, and do so plausibly.

If you can offer up a credible possible dissertation based on your ideas, then it follows that the dissertation you actually wind up writing will benefit from this early exercise. Your proposal will get finished faster, and so will your dissertation-because unlike diamonds, research proposals (and dissertations) are not forever. And doctoral study shouldn't be, either.

Master's thesis proposal: Natural Language Processing of Textual Use Cases

Advisor: Vladimir Mencl
Student: Jaroslav Dra┼żan

The design of a software system or component starts with specifying its requirements; traditionally, use cases written in natural language (English) are used for this task. Based on the simple and uniform sentence structure used in textual use cases [10], a conversion scheme [1, 3] has been proposed in the Procasor project [12] to derive behavior specifications from textual use cases. The scheme has been implemented in a prototype tool, employing a suite of readily available natural language processing tools [7, 8, 9].

In this preliminary work, certain issues remain open, such as evaluating the quality of the parse tree provided by the linguistic tools. Recent advances in the natural language processing tools [5] permit to obtain several possible parse trees for a sentence; furthermore, there are several different parsers available which may yield different parse trees.

The goal of the thesis is to build on the conversion scheme described in [1, 3] and propose metric to evaluate the quality of a parse tree. The thesis should address the issue of evaluating several parse trees of a sentence specifying a use case step and possibly also the issue of combining the information available in the parse trees. The thesis should also address the issue of constructing matching event tokens for complementary send / receive actions in use case models of communicating entities. The thesis should be supported by a proof-of-the-concept implementation.


[1] Mencl, V.: Deriving Behavior Specifications from Textual Use Cases, in Proceedings of Workshop on Intelligent Technologies for Software Engineering (WITSE04, Sep 21, 2004, part of ASE 2004), Linz, Austria, ISBN 3-85403-180-7, pp. 331-341, Oesterreichische Computer Gesellschaft, Sep 2004

[2] Plasil, F., Mencl, V.: Getting "Whole Picture" Behavior in a Use Case Model, in Transactions of the SDPS: Journal of Integrated Design and Process Science, vol. 7, no. 4, pp. 63-79, Dec 2003, ISSN-1092-0617, publisher: Society for Design and Process Science, Grandview, Texas, slightly modified version of paper published in Proceedings of IDPT 2003, Dec 2003

[3] Mencl, V.: Converting Textual Use Cases into Behavior Specifications, Tech. Report No. 2004/5, Dept. of SW Engineering, Charles University, Prague, Aug 2004

[4] Plasil, F., Mencl, V.: Use Cases: Assembling "Whole Picture" Behavior, Technical Report 02/11, Department of Computer Science, University of New Hampshire, NH, U.S.A., Nov 2002

[5] Bikel, D. M. : Design of a Multi-lingual, Parallel-processing Statistical Parsing Engine, in Proceedings of HLT 2002,

[6] Eugene Charniak: Statistical Techniques for Natural Language Parsing, AI Magazine 18(4): 33-44, 1997,

[7] Michael Collins: A New Statistical Parser Based on Bigram Lexical Dependencies., Proceedings of 34th Annual Meeting of the Association for Computational Linguistics, ACL 1996, 24-27 June 1996, University of California, Santa Cruz, California, USA, Morgan Kaufmann Publishers, 1996,

[8] Adwait Ratnaparkhi: A Maximum Entropy Part-Of-Speech Tagger, Proceedings of the Empirical Methods in Natural Language Processing Conference, May 17-18, 1996. University of Pennsylvania, 1996,

[9] Minnen, G., Carroll J., Pearce, D.: Applied morphological processing of English, Natural Language Engineering, 7(3), pp. 207-223, 2001,

[10] Cockburn, A.: Writing Effective Use Cases, Addison-Wesley Pub Co, ISBN: 0201702258, 1st edition, Jan 2000

[11] List of related work on requirement specifications and use cases,

[12] Procasor project,

One thought on “Adwait Ratnaparkhis Dissertation Proposal

Leave a Reply

Your email address will not be published. Required fields are marked *