Lecture 13 Revision Questions
300958 (COMP3020) Social Web Analytics
2014, Q6
2015, Q2
(a) Provide one reason for and one reason against using stop word removal when pre-
processing text for a search engine.
(b) Compute the cosine similarity score of each document to the query “T1 T3” by using term
frequencies.
(c) Which tweet is more similar to the query “T1 T3”? Justify your answer.
(d) A colleague tells you to include each word five times in the query to get better results
(e.g. search for T1 T1 T1 T1 T1 T3 T3 T3 T3 T3. Explain why this is a good or bad idea.
2014, Q1
Using this graph:
(a) Construct the adjacency matrix
(b) Tabulate the degree distribution.
(c) Calculate the graph density.
(d) Calculate the betweenness centrality for each vertex.
2013, Q3
2013, Q5

2014, Q8
2014, Q8
2017, Q6
The number of tweets containing “#jumpy” were counted for five days before and after the Jumpy advertising
campaign. A set of control tweets were also counted. The counts are listed below. We want to determine if the
advertising campaign had an effect of the number of tweets containing “#jumpy”.
(a) State the Null and Alternative hypothesis of the test that needs to be conducted.
(b) Estimate the effect caused by the change in company.
(c) State which test statistic should be used for the test and calculate the test statistics.
Control Impact
Before 91, 111, 100, 89, 88 171, 190, 194, 180, 193
After 69, 83, 71, 74, 88 104, 91, 95, 115, 111
2017, Q6
(d) Given the following randomisation distribution, what is the conclusion of the test?
2015, Q8
2015, Q8
2013, Q8
(a) Calculate and tabulate the word distribution for both positive
and negative sentiment.
(b) Compute the probability of the tweet “fun teeth shine” using
Naïve Bayes classification.
(c) Determine if the tweet “fun teeth shine” has positive or
negative sentiment based on the probabilities.
The Funfun Toothpaste company wants to gauge the sentiment towards
their toothpaste. Given the following set of labelled tweets:
Positive tweets:
• My teeth shine #funfun
• #funfun love my fun teeth
• #funfun is fun fun
Negative tweets:
• No shine #funfun
• No love fun fun
• Where is my teeth shine #funfun