This page gives some guidelines on how to find technical literature. It
is targeted to students supervised by myself,
but may be useful for a wider audience.
Note: This page contains links to commercial sites that are only
accessible from the utwente.nl domain because of campus licenses
bought by the University of Twente Library (other universities or individuals
may also have a license, of course).
Types of Literature
Almost any practical project at university starts with an orientation on
the topic of the project by studying the relevant literature. This literature
may consist of:
Books. In general, books present the more established part of a
theory. Because it may take quite some time to write a book and have it
published, the information in a book may not always present the state of
the art. There also exist books, however, that have had short production
times. Often, these books contain a collection of contributions from various
authors.
Scientific journal articles. The general statement to be made about
journals is that they contain specialized articles that are, in general,
more detailed than papers in conference proceedings. Articles are normally
only accepted for publication after a through peer review procedure (an
editor asks several experts in the field their opinions on a manuscript
and then takes a decision on acceptance or rejection; it is quite common
that reviewers ask to rewrite parts of the manuscript). The review and
publication procedure can be lengthy and it is not uncommon that two or
three years pass between the submission of a manuscript and its publication.
Papers in conference proceedings. Most conferences in my field are
annual or biannual events. Those interested to present their work, will
normally have to submit their contributions in advance for review. After
selection of the contributions, authors are asked to prepare the final
versions of their papers in time such that the conference proceedings
can be printed and distributed to the participants at the conference site.
The complete procedure takes less than a year and this means that conference
papers are the best source for the state of the art. They are, in general,
less detailed than journal articles as a consequence of the limit to the
number of pages imposed by the conference organization. Another disadvantage
of conference proceedings is that some of them are not always officially
published (they don't e.g. have an ISBN).
Ph.D. theses, M.Sc. theses, and other reports. The Ph.D. is the
highest academic degree and requires the writing of a thesis that is the
result of several years of work. It contains in general valuable albeit
very specialized information. Theses written to obtain an M.Sc. degree
normally have a different level (the work is done in much shorter time,
the student that wrote it was less experienced). Reports written by students
in other stages of their studies may also be relevant to your project.
Finally, so-called "internal reports", reports written by staff members
of some organization can contain useful information. There are several
reasons for writing internal reports: the necessity of documenting technical
work that is not scientifically important but essential in itself, confidentiality
of results, the wish to give some status to a manuscript between its moment
of submission to a conference or journal and its moment of acceptance (if
the submission is rejected, one has at least the internal report).
Web pages. Yet another source of information is formed by WWW pages
on the Internet. The quality of the information cannot be relied upon,
however, as anybody can publish anything on the web without having to pass
through a review process. Besides, information that was available today
may be gone tomorrow. In spite of this, if you are sufficiently critical
yourself, you may discover many valuable pages. There also exist on-line
scientific journals (without a printed version) that present material that
has passed a review process.
Other documentation. Other sources of documentation may be manuals
of the software or hardware to be used in your project. Popular-scientific
magazines or commercial technical magazines may be another source of useful
information.
Search procedures
Suppose that you want to collect literature about some topic. What can
you do?
Talk to your supervisors. They will probably have quite some expertise
in the field of your project and may know the relevant literature. In my
case, I have collected the literature that I have (partially) read in a
data
base that is accessible through the WWW. My database reflects my research
interests from the time that I started to work on my Ph.D. until now:
VLSI design, design automation, computational intelligence, signal
processing and telecommunication.
A good starting point is the home page
of the University
of Twente Library. You will find there the most complete information on
scientific
databases that are searchable on-line for students and staff of the
University of Twente. Many of them are only
accessible when you are using a computer in the University of Twente
domain (use a VPN connection when outside this domain).
The databases that I use most are:
ISI Web of Knowledge,
also known as the Science Citation Index.
It only contains journal articles (no conference papers).
Abstracts are sometimes available. Apart from the possibility to search
on title words, author, etc., it also has the possibility of the cited
reference search. The latter is quite interesting. Suppose that you
know of some key papers that are sufficiently old, then the cited reference
search will present you the articles that have cited those papers and may
report about progress on the specific topics. So, this search will lead
you forward in time.
Scopus, a large database of journal
and conference papers which also offers tracing citations.
Use the list of references of publications already available to you. Often,
you can judge from the way some publication has been cited whether the
publication is useful for you. Repeating this process will lead you backward
in time.
A powerful free on-line database that includes the tracing of citations
is Google Scholar.
A free on-line database that is quite interesting is
CiteSeer.
It is entirely based on publications of which
the full text is available on line. Searches based on the contents of
these articles as well as citation searches are provided. So, the
limitation of this database is that only documents whose full text is
(or has been) available on the WWW are included. This is at the same
time an advantage as the documents reported can directly be accessed.
Yet another on-line database is the
Collection of Computer Science Bibliographies
. It offers a search function on multiple bibliographic databases
(sometimes the same item is found in different databases and presented
multiple times).
Suppose that you have encountered a publication relevant to your research.
It may be worthwhile to see whether the authors have more relevant publications.
Many academic researchers have their list of publications on-line. Try
to locate these lists by using a search engine such as
Google.
Obtaining the literature
Suppose that you have identified the items that you want to read, how do
you obtain them?
A good starting point is again
the home page of the University
of Twente Library.
You can search the catalog to see whether a book is physically present
in the collection. You can also see which electronic subscriptions
exist on a jounal-by-journal basis. There are a few bulk subscriptions
giving access to most publications in our field:
IEEE Xplore for
publications by the Institute of Electrical and Electronic
Engineers, both journal papers (such as IEEE Transactions on ...)
and conference contributions.
ACM Digital Library for
journal and conference publications of the Association of Computing
Machinery.
Springerlink for
publications by publisher Springer; note that many conference
proceedings are published as an issue of the Springer Lecture Notes
on Computer Science (LNCS), which is officially a journal.
An alternative way to obtain the item you want is to check the web sites
of the authors (see above on how to locate them). Even when the document
you are looking for is not on-line, you may consider to request for a preprint
of the item by email. However, do not disturb people unnecessarily. Authors'
pages are especially interesting for obtaining Ph.D. theses, M.Sc. theses
and internal reports that are more difficult to access through the library.