IR Model: Boolean Model

1
IR Model Boolean Model

• Dr. Ying Xie

2
Web Search Process
Crawler
Information Need
Formulation
Indexing
Query Rep
Inverted index and web graph
Ranking
Ranked List
Learning
User Relevance Feedback
3

• Boolean Retrieval Model

4
Definition

• What is Boolean Model for IR?
• - First, by Boolean Model, the query is
  expressed as a boolean expression.
• e.g. information and ((retrieval or search) not
  filtering)
• - The retrieval is the process of finding
  all the documents that satisfy the query (boolean
  expression).

5
information and ((retrieval or search) not
filtering)
6
information and ((retrieval or search) not
filtering)
7
information and ((retrieval or search) not
filtering)
8
R a document is predicted as relevant to a
query expression iff it satisfies the query
expression information and
retrieval Each query term specifies a set of
documents That satisfy the term
2
4
8
16
32
64
information
2
3
8
21
32
1
retrieval
2
8
32
information and retrieval
9
information or retrieval
2
4
8
16
32
64
information
2
3
8
21
32
1
retrieval
21
32
64
2
3
4
8
1
16
information or retrieval
10
information not retrieval
2
4
8
16
32
64
information
2
3
8
21
32
1
retrieval
4
16
64
information not retrieval
11
information and ((retrieval or search) not
filtering)
Information 2,4,8,16,32,64 Retrieval
1,2,3,8,21,32 Search 1,2,5,8,21 Filterin
g 8,32,64
12
Disadvantages

• There is no ranking among the retrieved documents
  
• All terms are equally important (not really true)
• Query language is complicated