PhD Guidelines

1. Research (Date: 1577)

1. Careful or diligent search
2. Studious inquiry or examination; especially: investigation or experimentation aimed at the discovery and interpretation of facts, revision of accepted theories or laws in the light of new facts, or practical application of such new or revised theories or laws
3. The collecting of information about a particular subject

Good research follows a pattern (adapted from “The Craft of Research”, p.49):

• practical problem: problem that you experience or you observe in the “reality” and which manifests itself as a cost in time, satisfaction, money, etc… It is important to motivate why the problem is important enough to be worth the research. The practical problem is associated with a topic, which is the area in which the research will be done.
• state of the art analysis: once a practical problem is identified, a state of the art analysis is done to identify and evaluate all the existing solutions to the practical problem. The state of the art analysis includes a literature analysis (i.e. review of the published research results) and a best practice analysis (i.e. review of the current industrial practice). At this point, either the practical problem is solved (still be worth to write a technical report on the results) or none of the existing solutions are satisfactory and you can carry the research.
• research problem: practical problem reformulated by the researcher in a way which states how the current state of the art presents an incomplete or flawed understanding.
• research solution: solution to the research problem, which could be applied to solve the practical problem. Usually, a hypothesis is stated and is then validated using some methods. It is important to show how the research solution contributes to solve the practical problem. Note that hypothesis which have been proven invalid might also be published.

We recommend reading “The Craft of Research”.
References:

The Craft of Research
Wayne C. Booth, Gregory G. Colomb, Joseph M. Williams
The University of Chicago Press
ISBN 0-226-06583-9

• make sure that an existing research solution is not reinvented by ignorance of its existence,
• demonstrate the logical continuity between previous and present work (APA Publication Manual, p. 11).

State of the art analysis is an important factor differentiating research done in academy versus research done in industry. In the industry, it is enough to have a good enough understanding of what is going on in the industry to develop a competitive and original product. In academic research, it is essential to have a complete understanding of what is relevant to the research in order to have an original and innovative contribution.

In the state of the art analysis, your goal is to collect and archive important information. Information includes:

• who are the key players in your topic
• what are the key conferences and journals in your topic
• which are the seminal papers in your topic

These data have tremendous value and are worth sharing with colleagues. To do so we strongly recommend using Endnote as software to manage your state of the art.

Information sources (non exhaustive list):

Digital Libraries

Reference Web Sites

Wiley:
http://www.interscience.wiley.com/

Professional Associations and Research Groups:

Dictionaries and Handbooks:

Stanford Philosophy Encyclopedia:
http://plato.stanford.edu/

Principa Cybernetica Web:
http://pespmc1.vub.ac.be/

The MIT Encyclopedia of Cognitive Sciences:
http://cognet.mit.edu/erefs/mit-encyclopedia-of-cognitive-sciences-mitecs

Encyclopedia Universalis:
Access yet to be defined.

Formal Models and Semantics – Volume B
Handbook of Theoretical Computer Science
isbn 0-444-88074-7
McDermid

Software Engineers Reference Book
Butterworth Heinemann
isbn 0-7506-0813-7

General Purpose Dictionaires:
http://www.dictionary.com/
http://www.m-w.com/ or http://www.webster.com/

Journal Web Sites:

Newsgroup:

National Library, Nebis, EPFL, I&C, LAMS Libraries:

EPFL Central Library:
http://infoscience.epfl.ch/

Tools:

• to help remembering
• o help understanding
• to help gaining perspective
• to be known in the field of research

Publishing is also useful for the lab as it allows to be known in the field, and so to get research grants and, ultimately, to pay you (or your facilities) ! As publishing is essential to be known, it is important to carefully select where to publish.

Where to publish ?

First you have to identify the conferences, special issues of journal or journals that you want to target. To help you, people compile lists.

Then you need to assess the quality of the conference / publication. Conferences and journals are ranked. At least two ranking exists. One is the citation index. It is sold by www.isinet.com and applies only to journal.

How to publish ?

It is important to read a lot of journal and conference papers to understand how to structure them and what is the state of the art. In addition, advices and given in the next section.

A lot of the work in writing a paper is to make sure you show an in-depth understanding of the field. A very good idea is to have a target paper as an example of what needs to be written. For example, the following journal paper gives a good example of a paper anchored in practice with a solid theoretical part.

What to Publish ?

You can find a lot of guidelines on what to publish. The first book (even if not in the field of CS) gives a good overview of what should be in a paper. A lot of web site also gives such info.

References:

Publication Manual of the American Psychologocial Association
American Psychological Association
isbn 1-55798-241-4

Writing Successfuly in Science
Maeve O’Connor
isbn 0-412-44630-8

How to Publish a Scientific Paper
Robert Day
isbn 0-521-36572-4

Webography:
http://www.cs.columbia.edu/~hgs/etc/writing-style.html
http://www.cs.cmu.edu/afs/cs.cmu.edu/user/mleone/web/how-to.html

From (SIGCHI) CHI Conference:
http://www.acm.org/sigchi/chi97/call/papers/info.html

FROM (SIGOPS) SOSP Conference:
http://www.usenix.org/events/samples/submit/advice.html

From (SIGCOMM) SIGCOMM Conference:
http://www.sigcomm.org/

…

In addition, you will have the following milestones to achieve:

• get a conference paper accepted in first year
• get a journal paper submitted in second year
• get a journal paper accepted + second journal paper submitted in third year

These requirements are known as the “PhD Process”. They are here to make sure you gain real benefits (i.e. personal visibility) from your research. Without this, your research is far less beneficial for you.

Examples of possible target publication are:

• Magazines:
  • Communications of the ACM
  • IEEE Software
  • …
• Journals:
  • ACM Computing Surveys
  • IEEE Transaction on Software Engineering
  • ACM Transactions on Information Systems
  • Journal of Computer and System Sciense
  • Formal Methods in System Design
  • ACM Transactions on Systems, Man and Cybernetics – part B
  • IEEE Transaction on Knowledge and Data Engineering
  • Software Practice and Experience
  • IEEE Transaction on Education
  • Data and Knowledge Engineering
  • …
• Conferences:
  • EDOC 200x submission in March (conference in September)
  • UML 200x submission in March (conference in October)
  • ICSE 200x submission in August (conference in May)
  • ECOOP 200x submission in October (conference in June)
  • OOPSLA 200x submission in March (conference in October)
  • …