Week 4 slides

General Writing - Audience
What is their level of knowledge?
Advanced, intermediate, basic?
Hard to start too basic – but have to use the right
terminology (use a sample paper)
What is their goal in reading the paper?
General information – survey
Specific results
What is your goal in writing the paper?
Report specific result(s)
General info
Rebuttal
Other
General Writing - Audience
Appropriate title
Do you want to appeal to a broad audience
or target a specific audience?
“Graph Structure in the Web”
“Learning representations by back-propagating
errors”
“Removal Bias: A New Cause of Code Growth in
Genetic Programming”
“The Task of the Referee”
General Writing - Clarity
Language - Be precise
Avoid jargon (as appropriate)
Use definitions where necessary – avoid too
many new definitions
Be careful with abbreviations
English – Be concise
Active sentences
Simple Sentences
General Writing - Precise
•The algorithm is slow. How slow?
•This algorithm has a high computational complexity.
How high?
•Large instances of the problem are computationally
infeasible. What is large? What is infeasible?
•The results show that algorithm A is better than
algorithm B. Better meaning what? Faster? Uses less
memory? More accurate? Easier to understand?
Easier to use?
•First, the objects are sorted. Sorted by what?
General Writing – Clarity
Be careful when using pronouns – “it”
“Algorithm A is 50% more efficient than algorithm
X. It uses a bubble sort to improve efficiency,
although a quick sort would be better.”
General Writing – Active
“Protecting the confidentiality of information
manipulated by computing systems is a longstanding yet increasingly important question.”
“A longstanding question of importance is how
to protect the confidentiality of information that is
being manipulated by computing systems.”
“An important, longstanding question is how to
protect the confidentiality of information that is
manipulated by computing systems.”
Which is easier to read?
General Writing – Clear, Active
“The research process involved running
each of the three algorithms on each of five
separate data sets.”
“All of the algorithms were applied to a
collection of five data sets.”
“We ran all three algorithms on five
separate data sets.”
General Writing – Active
By running the strongly connected component
algorithm, we find that there is a single large SCC
consisting of about 56 million pages, all other
components are significantly smaller in size. This
amounts to barely 28% of all the pages in our crawl.
One may now ask: where have all the other pages
gone? The answer to this question reveals some
fascinating detailed structure in the Web graph; to
expose this and to further study the issues of the
diameter and average distance, we conducted a
further series of experiments.