Description
Q1. Here is the link to the list of UFO sightings : http://www.nuforc.org/
webreports/ndxshape.html. We are interested in UFOs corresponding to one
of three shapes: circle, triangle and fireball.
1. This is a structured but dirty dataset. You should collect all sightings from
that list up to and including September 9, 2013. Specifically, you should
represent each sighting by these eight features: Data of sighting, Time of
sighting, City, State, Shape, Duration, Summary, Posted date (when the
sighting was posted to the website). You should do your best to convert
all Durations to seconds, whenever possible. Turn in your cleaned dataset
in csv format. Keep in mind a few guidelines:
• If a duration has a“<” sign, you should simply ignore the“<” sign.
For example if the duration is specified as “< 1 minute”, consider
the duration to be “1 minute”. You should subsequently convert “1
minute” to “60 seconds”.
• If a duration has a range, use the upper limit as its value. For example, if the duration is listed as “5-8 minutes”, you should consider the
duration as “8 minutes”. (Again, you will need to eventually convert
minutes into seconds).
• You may encounter some other oddities in the data. Do your best
to extract maximum value from the messy data; be sure to explain
to us the decisions you have made in terms of data extraction and
cleaning.
2. Based on your cleaned data, answer the following questions using SQL
queries :
• How many UFO sightings in Alaska?
• How many UFO sightings of durations less than 2 minutes in NY?
• What’s the average duration of UFO sightings of fireball?
1
• Which year has the maximum number of sightings?
• During the December, 2012, how many states witnessed UFO more
than 30 times?
Q2. Download the first dataset from http://acube.di.unipi.it/tmn-dataset/
and structure it. The final structured dataset is expected to be a matrix with
five columns corresponding to date (in the format mmddyyyy), source, category,
the number of words (not including punctuation) in the title, the number of
commas in the description. Based on the structured dataset you get, answer
the following answers:
• Which category of news is most popular?
• How many business events happened during April to May 2011?
• How many news events come from usatoday.com?
• Which newspaper website has the longest title in average?
• Which newspaper website uses least commas (in average) in description?
Q3. Pick one dataset from the list here http://www.infochimps.com/datasets
and join it with the UFO dataset (joining on date or location is one place to
start). Answer the following questions :
• Describe the dataset and why you chose it
• Create 5 questions based on the joined data set and answer them.
• Create one interesting visualization and explain what it shows
Q4. Here are two readings :
• http://bits.blogs.nytimes.com/2013/06/01/why-big-data-is-not-truth/
?_r=1&
• http://vldb.org/pvldb/vol5/p1674_tamraparnidasu_vldb2012.pdf
Both these articles concern our attempts as data scientists to get to some notion
of the ground truth. Please write 2-3 paragraphs with your thoughts after
reading them, and connect your ideas to any specific examples from your field
of study.
