02 Introduction to retrieval system

Question 1

what is information retrieval

Answer 1

science of search engine
- effectively get the right information to user
- efficiently get information to user
- relevance to see if the query and document is about the same topic

Question 2

general definition of IR

Answer 2

retrieval of relevant information from data sources which were not originally intended for access

Question 3

documents vs database records**

Answer 3

database records store in tuples, but how to match with textual records
retrieving:
- structured data (DB)
- free text (IR)

queries
- formally defined (DB)
- vague, imprecise (IR)

results
- exact, always correct (DB)
- sometime relevant (IR)

Question 4

imprecision in IR

Answer 4

most algorithms in computer science have ‘right’ answers
IR techniques are essentially heuristic as we do not know the right answers

Question 5

users classification

Answer 5

naive to expert

1. professionals
   - ‘accommodation in glasgow’
   - direct term search
2. general users
   - how do i get to …
3. laymen
   - doctor said i have… where do i get more information
   - long texts

Question 6

information pyramid

Answer 6

data
information = data in context
knowledge = basis for making decisions
wisdom

Question 7

what makes a document relevant

Answer 7

• does it contain all query terms
• contain many times
• fresh
• authoritative (have many links)
• doesnt contain too many ads
• doesnt contain spam
• has been clicked the most

Question 8

search engines

Answer 8

web search process is not a one search approach
- question answer assessment
- user engagement