Information Technology Reference
In-Depth Information
Avoid nonsense, absurdity, and over-generalization.
Execution was almost instantaneous.
The Web is infinitely large.
There is no limit to the possible efficiency gains.
Reference and Citation
You need to explain the relationship of your new work to existing work, showing
how your work builds on previous knowledge and how it differs from contributions
in other, relevant papers. The existing work is identified by reference to published
theses, articles, and books. All papers include a bibliography, that is, a list of such
references in a standardized format, and embedded in each paper's text there are
citations to the publications.
References, and discussion of them, serve three main purposes. They help demon-
strate that work is new: claims of originality are much more convincing in the context
of references to existing work that (from the reader's perspective) appears to be sim-
ilar. They demonstrate your knowledge of the research area, which helps the reader
to judge whether your statements are reliable. And they are pointers to background
reading.
Before including a reference, consider whether it will be of service to the reader. A
reference should be relevant, up-to-date, and reasonably accessible; and it should be
necessary. Don't add citations just to pad the bibliography. Refer to an original paper
in preference to a secondary source; to well-written material in preference to bad;
to a book, conference paper, or journal article in preference to a workshop paper;
to a workshop paper in preference to a manuscript (which have the disadvantage
of being unrefereed); and to formally published documents rather than Web pages.
Avoid reference to private communications and information provided in forums such
as seminars or talks—such information cannot be accessed or verified by the reader.
In the rare circumstance that you must refer to such material, do so via a footnote,
parenthetical remark, or acknowledgement.
If you discuss a paper or note some particular contribution it makes, it must be
cited. Otherwise, consider whether a reader needs the paper for knowledge in addition
to that in the other papers you cite. If the answer is no, perhaps it should be omitted.
At the same time, ensure that it is clear to the reader that you know all the pertinent
background literature.
Don't cite to support common knowledge. For example, use of a binary tree in an
algorithm doesn't require a reference to a data structures text. But claims, statements
of fact, and discussion of previous work should be substantiated by reference if not
substantiated within your write-up. This rule even applies to minor points. For some
readers the minor points could be of major interest.
In many papers, some of the references are to previous publications by the
same author. Such references establish the author's credentials as someone who
