25
Search Result Interface Hongning Wang CS@UVa

Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

Embed Size (px)

Citation preview

Page 1: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

Search Result Interface

Hongning WangCS@UVa

Page 2: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 2

Abstraction of search engine architecture

User

RankerIndexer

Doc Analyzer

Index results

Crawler

Doc Representation Query Rep

(Query)

EvaluationFeedback

CS@UVa

Indexed corpus

Ranking procedure

Page 3: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 3

Search interface

• Evolution of Google’s result interface– http

://www.seosmarty.com/google-interface-evolution/

CS@UVa

Page 4: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 4

Google’s

CS@UVa

Page 5: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 5

Bing’s

CS@UVa

Page 6: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 6

Yahoo’s

CS@UVa

Page 7: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS4501: Information Retrieval 7

Recap: Index compression

• Observation of posting files– Instead of storing docID in posting, we store gap

between docIDs, since they are ordered– Zipf’s law again: • The more frequent a word is, the smaller the gaps are• The less frequent a word is, the shorter the posting list

is

– Heavily biased distribution gives us great opportunity of compression!

Information theory: entropy measures compression difficulty.

CS@UVa

Page 8: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS4501: Information Retrieval 8

Recap: Search within inverted index

• Example: AND operation

Term1

Term2

1282 4 8 16 32 64

341 2 3 5 8 13

21

scan the postings

Time complexity:

Trick for speed-up: when performing multi-way join, starts from lowest frequency term to highest frequency ones

CS@UVa

Page 9: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS4501: Information Retrieval 9

Recap: Phrase query

• Generalized postings matching– Equality condition check with requirement of

position pattern between two query terms• e.g., T2.pos-T1.pos = 1 (T1 must be immediately before

T2 in any matched document)

– Proximity query: |T2.pos-T1.pos| ≤ k

128

34

2 4 8 16 32 64

1 2 3 5 8 13

21

Term1

Term2

scan the postings

CS@UVa

Page 10: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 10

What are there

• A list of links to the search result page– Text summarization of the retrieved document

• Title + text snippet

• Search suggestions– Related search– Spelling correction– Query auto-completion

• Vertical search– Image, shopping, news

• Knowledge graph– As a result of NLP techniques

It has been there since the search engine was born

Simple Q&A

CS@UVa

Page 11: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 11

Query auto-completion

CS@UVa

Page 12: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 12

Direct answers

• Advanced version of “I’m feeling lucky”

CS@UVa

Page 13: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 13

Experimental features

• Search result feedback

CS@UVa

Page 14: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 14

Experimental features

• Collaborative ranking

CS@UVa

Page 15: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 15

Experimental features

• Social panel

CS@UVa

Page 16: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 16

Instant search

CS@UVa

Page 17: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 17

Carrot2’s folder display

Organized results

CS@UVa

Page 18: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 18

Carrot2’s circle display

CS@UVa

Page 19: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 19

Carrot2’s foam tree display

CS@UVa

Page 20: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 20

BaiGoogledu

CS@UVa

Meta search engine

Page 21: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 21

PubMed

CS@UVa

Page 22: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 22

Considerations in result display

• Relevance– Order the results by relevance

• Diversity– Maximize the topical coverage of the displayed

results• Navigation– Help users easily explore the related search space• Query suggestion• Search by example

CS@UVa

Page 23: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 23

• In Human-Computer Interaction– Eye/Mouse tracking study of interaction between

users and search result page– Psychological study of user behaviors• Facet categories, text summaries, colors, positions

• In Information Retrieval– Less attention has been put in this aspect in

history– Attracting more and more research focus now

Research progress

CS@UVa

Page 24: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 24

Search result display in mobile device

• Unique characteristics of mobile device– Small screen size, limited bandwidth, input, data-

access and computation power– Multi-touch screen– Rich search context– Opportunities?

CS@UVa

Page 25: Search Result Interface Hongning Wang CS@UVa. Abstraction of search engine architecture User Ranker Indexer Doc Analyzer Index results Crawler Doc Representation

CS 4501: Information Retrieval 25

What you should know

• General considerations in search result display• Challenges and opportunities in mobile device

search result display

CS@UVa