THEORY ASSIGNMENTS, DEC-2022

Information Retrieval (CSE 4053)

Programme: B. Tech (CSE) Semester:7th

Full Marks: 20 Submission Dt.- 24. 12. 2022

Subject/Course Learning Outcome *Taxonomy Ques. Marks

Level Nos.
Outline the concepts and apply the basics of
indexing and querying of an information L1,L2,L3
1 2
retrieval system

Understand the data corpus used in

information retrieval systems. L1,L2,L3 2 2

Illustrate various components and experiment

with different compression techniques to
compress the index of dictionary and its L1,L2,L3 3,4 4
postings lists.

Apply retrieval models to construct information

retrieval systems L3,L4,L5 5, 6, 7, 8 8

Apply text clustering and classification

techniques for information retrieval. L3,L4,L5 9, 10 4

*Bloom’s taxonomy levels: Knowledge (L1), Comprehension (L2), Application (L3), Analysis (L4), Evaluation
(L5), Creation (L6)

Answer all questions. Each question carries equal mark.

 Assignment scores/markings depend on neatness, clarity and date of submission.

 Write your answers with enough detail about your approach and concepts used, so
that the grader will be able to understand it easily.
 You are allowed to use only those concepts which are covered in the lecture class till
date.

1. For a 2 word query, the postings lists are as mentioned below: [2, 5, 7, 13, 16, 18, 23, 32,
39, 56, 59, 63, 68, 76, 80] and [12, 56]. Show how many comparisons (step by step
representation) would be done to find out the intersection of the above two postings lists
using skip pointers with a skip length of P .
2. For a given document stored in the data warehouse, compress the words by applying
following preprocessing technique separately.
i. Normalization
ii. Stemming (Use Porter stemmer)
iii. Stop words removal
 Information retrieval is the activity of obtaining
information resources relevant to an information need from
a collection of information resources. Searches can be
based on full text or other content based indexing.
Automated information retrieval systems are used to reduce
what has been called "information overload". Many
universities and public libraries use IR systems to provide
access to books, journals and other documents. Web search
engines are the most visible IR applications.

3. Edit distance can be used for spelling correction in search queries.

(i) Define edit distance.
(ii) As an example of how to calculate edit distance efficiently, show how dynamic
programming can be used to calculate the edit distance between “system” and “item”.
4. Suppose the vocabulary for your inverted index consists of the following 6 terms:
elite, elope, ellipse, eloquent, eligible, elongate
Assume that the dictionary data structure used for this index stores the actual terms using
Dictionary-as-a-string storage with front coding and a block size of 3. Show the resulting
storage of the above vocabulary of 6 terms.
5. Consider a collection made of the 4 following documents (one document per line) :

D1: John gives a book to Mary

D2: John who reads a book loves Mary
D3: who does John think Mary love?
D4: John thinks a book is a good gift
- These documents are pre-processed using a stop-list and a stemmer. The resulting
index is built to allow by applying vector-based queries. Give a (graphical or textual)
representation of this index.
- We now focus on 3 terms belonging to the dictionary, namely book, love and Mary.
Compute the tf − idf -based vector representation for the 4 documents in the
collection.
- Consider the query “love Mary”. Give the results of a ranked retrieval for this query.
What document is (are) considered to be the most relevant ?
6. Compute the Jaccard matching score and the tf matching score for the following query-
document pairs.

q: [information on cars] d: “all you’ve ever wanted to know about cars”

q: [information on cars] d: “information on trucks, information on planes,
information on trains”
q: [red cars and red trucks] d: “cops stop red cars more often”
7. The figure below shows the output of an information retrieval system on two queries. Crosses
correspond to the relevant documents, dashes to non-relevant documents. Let the two documents
contain 3 and 6 relevant documents, respectively, but only those shown in the figure are retrieved
by the system, not the others.

Rank Q1 Q2
1 X -
2 - X
3 - -
4 X X
5 X -
6 - X
7 - X
8 - -
9 - X
10 - X

(a) Draw the precision-recall curve

(b) Compute the Mean Average Precision
(c) Compute the R-precision
8. Rank the documents in collection {d1 , d2 } for query q using the language model approach
to IR introduced in class with Jelinek-Mercer smoothing. Use the mixture coefficient λ =
0.4.
d1 : Scottish sheep getting smaller due to climate change study says
d2 : The analysis has shown a dramatic shift in the natural ranges for US
Bird species in response to climate change
Query q: climate change
9. Classify whether a given person is a male or a female based on the measured features.
The features include height, weight, and foot size. Data set is shown below
sex height (feet) weight (lbs) foot size
male 6 180 12
(inches)
male 5.92 (5'11") 190 11
male 5.58 (5'7") 170 12
male 5.92 (5'11") 165 10
female 5 100 6
female 5.5 (5'6") 150 8
female 5.42 (5'5") 130 7
female 5.75 (5'9") 150 9

Below is a sample to be classified using Neive Bayes algorithm as a male or female.

sex height weight foot size

sample 6(feet) 130
(lbs) 8(inches)

10. Based on the data below, estimate a multinomial Naive Bayes classifier and apply the
classifier to the test document. Calculate the probability that the classifier assigns the test
document to c = China or.

docID words in document In c= China?

training set 1 Taipei Taiwan yes
2 Macao Taiwan Shanghai yes
3 Japan Sapporo no
4 Sapporo Osaka Taiwan no
5 London no
test set 6 Taiwan Taiwan Sapporo ?