applying the user over ranking hypothesis to query
play

Applying the User-over-Ranking Hypothesis to Query Formulation - PowerPoint PPT Presentation

Aug 24, 2023 •304 likes •1.08k views

Applying the User-over-Ranking Hypothesis to Query Formulation Matthias Hagen Benno Stein Bauhaus-Universit at Weimar matthias.hagen@uni-weimar.de ICTIR 2011 Bertinoro, Italy September 14, 2011 Matthias Hagen, Benno Stein Applying the

  1. Applying the User-over-Ranking Hypothesis to Query Formulation Matthias Hagen Benno Stein Bauhaus-Universit¨ at Weimar matthias.hagen@uni-weimar.de ICTIR 2011 Bertinoro, Italy September 14, 2011 Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 1

  2. What is the User-over-Ranking hypothesis? Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 2

  3. The User-over-Ranking Hypothesis [Stein and Hagen, ECIR 2011] Queries returning as many results as the user can consider increase retrieval performance. Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 3

  4. The User-over-Ranking Hypothesis [Stein and Hagen, ECIR 2011] Queries returning as many results as the user can consider increase retrieval performance. Fine print: If ranking works: great! Use case is not some query like ebay . But more involved information needs, automatic systems, etc. Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 3

  5. Assumption 1: More keywords = more specific Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 4

  6. Assumption 1: More keywords = more specific Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 4

  7. Assumption 1: More keywords = more specific 10 6 underspecific Result list length overspecific 0 Query 1 10 20 length Specificity of Queries Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 5

  8. Assumption 2: User can arbitrarily specify information need 10 6 underspecific Result list length overspecific 0 Query 1 10 20 length Specificity of Queries Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 6

  9. Assumption 2: User can arbitrarily specify information need 10 6 underspecific Result list length overspecific 0 Query 1 10 20 length Specificity of Queries Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 6

  10. Assumption 3: User can consider about k results. 10 6 underspecific Result list length k overspecific 0 Query Result 10 3 10 4 1 10 20 0 10 100 length list length Processing k capacity Specificity of Queries Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 7

  11. Hypothesis: Specificity matches k = Optimum retrieval optimum retrieval 10 6 underspecific ⇔ Result result set size = k list length k overspecific 0 Query Result 10 3 10 4 1 10 20 0 10 100 length list length Processing k capacity Specificity of Queries Probability for Retrieval Success Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 8

  12. What is this hypothesis good for? Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 9

  13. Query Formulation Scenario Given a set W of keywords Find a good query Q ⊆ W Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 10

  14. Query Formulation Scenario Given a set W of keywords Find a good query Q ⊆ W Previous approach [Lee et al., CIKM 2009] Learnt ranking function identifies the m best keywords from W . Based on: Known relevant documents Unrestricted index access Manually tuned m for each set W Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 10

  15. Consider for instance . . . Known-Item Finding Scenario User accessed a document But did not store it Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 11

  16. Consider for instance . . . Known-Item Finding Scenario User accessed a document How can she find it again? But did not store it Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 11

  17. Consider for instance . . . Known-Item Finding Scenario User accessed a document How can she find it again? But did not store it Solution Remember some keywords information retrieval query formulation web search search session user support search engine cost optimization Query a search engine Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 11

  18. But what query to formulate with the keywords? Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 12

  19. Single keywords? information retrieval / query / / / / / / / / / / formulation / / / / / / / / / / / / / / / / / / web / / / / / / / search / / / / / / / / / / search / / / / / / / / / / / / session / / / / / / / / / / / / user/ / / / / / / / support / / / / / / / / / / / / search / / / / / / / / / / / / engine / / / / / / / / / / cost/ / / / / / / / optimization / / / / / / / / / / / / / / / / / / / / Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 13

  20. Single keywords? / information/ / / / / / / / / / / / / / / / / / / / retrieval / / / / / / / / / / / / / / query formulation / web / / / / / / / search / / / / / / / / / / search / / / / / / / / / / / / session / / / / / / / / / / / / user/ / / / / / / / support / / / / / / / / / / / / / search / / / / / / / / / / / engine / / / / / / / / / / cost/ / / / / / / / optimization / / / / / / / / / / / / / / / / / / / / Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 13

  21. Single keywords? / information/ / / / / / / / / / / / / / / / / / / / retrieval / / / / / / / / / / / / / / / query / / / / / / / / / / formulation / / / / / / / / / / / / / / / / / web search / search / / / / / / / / / / / / session / / / / / / / / / / / / user/ / / / / / / / support / / / / / / / / / / / / search / / / / / / / / / / / / engine / / / / / / / / / / cost/ / / / / / / / optimization / / / / / / / / / / / / / / / / / / / / Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 13

  22. Single keywords? Underspecific! / information/ / / / / / / / / / / / / / / / / / / / retrieval / / / / / / / / / / / / / / / query / / / / / / / / / / formulation / / / / / / / / / / / / / / / / / web search / search / / / / / / / / / / / / session / / / / / / / / / / / / user/ / / / / / / / support / / / / / / / / / / / / search / / / / / / / / / / / / engine / / / / / / / / / / cost/ / / / / / / / optimization / / / / / / / / / / / / / / / / / / / / Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 13

  23. All keywords at once? information retrieval query formulation web search search session user support search engine cost optimization Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 13

  24. All keywords at once? Overspecific! information retrieval query formulation web search search session user support search engine cost optimization Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 13

  25. Remember the hypothesis . . . not too many results! Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 14

  26. Solution: As many keywords as possible! information retrieval query formulation web search search session / user/ / / / / / / / support / / / / / / / / / / / search engine cost optimization Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 15

  27. “As many keywords as possible”-Query Characteristics Captures most of the remembered keywords Best possible description of the known-item Not too many results user can check complete list → Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 16

  28. “As many keywords as possible”-Query Characteristics Captures most of the remembered keywords Best possible description of the known-item Not too many results user can check complete list → Problem Relevant documents not known No web index at user site Query size not known Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 16

  29. “As many keywords as possible”-Query Characteristics Captures most of the remembered keywords Best possible description of the known-item Not too many results user can check complete list → Problem Relevant documents not known No web index at user site Lee et al. not applicable → Query size not known Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 16

  30. We propose an approach for this scenario . . . Matthias Hagen, Benno Stein Applying the User-over-Ranking Hypothesis to Query Formulation 17

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

Ranking linked data Web graph, PageRank, Topic-specific PageRank and HITS Web Search Overview

Ranking linked data Web graph, PageRank, Topic-specific PageRank and HITS Web Search Overview

Ranking linked data Web graph, PageRank, Topic-specific PageRank and HITS Web Search Overview Indexes Query Indexi xing Ranki king Applica cation Results Documents User Information Query y Query analys ysis proce cess ssing

387 views • 35 slides
Ranking linked data Web graph, PageRank, Topic-specific PageRank and HITS Web Search 1 Overview

Ranking linked data Web graph, PageRank, Topic-specific PageRank and HITS Web Search 1 Overview

Ranking linked data Web graph, PageRank, Topic-specific PageRank and HITS Web Search 1 Overview Indexes Query Indexi xing Ranki king Applica cation Results Documents User Information Query y Query analys ysis proce cess ssing

848 views • 37 slides
Depth Estimation for Ranking Query Optimization KarlSchnaitter,UCSantaCruz

Depth Estimation for Ranking Query Optimization KarlSchnaitter,UCSantaCruz

Depth Estimation for Ranking Query Optimization KarlSchnaitter,UCSantaCruz JoshuaSpiegel,BEASystems,Inc. NeoklisPolyzotis,UCSantaCruz Relational Ranking Queries SELECT h.hid, r.rid, e.eid FROM Hotels h,

856 views • 33 slides
Efficient query processing Efficient scoring, distributed query processing Web Search 1 Ranking

Efficient query processing Efficient scoring, distributed query processing Web Search 1 Ranking

Efficient query processing Efficient scoring, distributed query processing Web Search 1 Ranking functions In general, document scoring functions are of the form The BM25 function, is one of the best performing: The term frequency is

648 views • 31 slides
Query Independent Scholarly Article Ranking Shuai Ma, Chen Gong , Renjun Hu, Dongsheng Luo,

Query Independent Scholarly Article Ranking Shuai Ma, Chen Gong , Renjun Hu, Dongsheng Luo,

Query Independent Scholarly Article Ranking Shuai Ma, Chen Gong , Renjun Hu, Dongsheng Luo, Chunming Hu, Jinpeng Huai SKLSDE Lab, Beihang University, China Beijing Advanced Innovation Center for Big Data and Brain Computing Query Independent

597 views • 34 slides
Query Processing Relevance feedback; query expansion; Web Search 1 Overview Indexes Query

Query Processing Relevance feedback; query expansion; Web Search 1 Overview Indexes Query

Query Processing Relevance feedback; query expansion; Web Search 1 Overview Indexes Query Indexi xing Ranki king Applica cation Results Documents User Information Query y Query analys ysis proce cess ssing Multimedia

740 views • 32 slides
Applying Hypothesis-Testing Methods to Help Inform Causality Conclusions from Epidemiology

Applying Hypothesis-Testing Methods to Help Inform Causality Conclusions from Epidemiology

Applying Hypothesis-Testing Methods to Help Inform Causality Conclusions from Epidemiology Studies SABINE LANGE, PHD, DABT LALITA SHRESTHA, PHD FEBRUARY 19, 2020 TOXICOLOGY, RISK ASSESSMENT, AND RESEARCH DIVISION TEXAS COMMISSION ON

832 views • 55 slides
Ranked Query Processing: Relevance a) Order-based Paradigm Preference Q Kevin Chen-Chuan

Ranked Query Processing: Relevance a) Order-based Paradigm Preference Q Kevin Chen-Chuan

Ranking Ordering according to the degree of some fuzzy notions: Similarity (or dissimilarity) Ranked Query Processing: Relevance a) Order-based Paradigm Preference Q Kevin Chen-Chuan Chang ranking 2 Query models for

295 views • 13 slides
III.5 Advanced Query Types (MRS book, Chapters 9+10; Baeza-Yates, Chapters 5+13) 5.1 Query

III.5 Advanced Query Types (MRS book, Chapters 9+10; Baeza-Yates, Chapters 5+13) 5.1 Query

III.5 Advanced Query Types (MRS book, Chapters 9+10; Baeza-Yates, Chapters 5+13) 5.1 Query Expansion & Relevance Feedback 5.2 Vague Search: Phrases, Proximity-based Ranking, More Similarity Measures: Phonetic, Editex, Soundex 5.3

824 views • 55 slides
CS6200 Information Retrieval Jesse Anderton College of Computer and Information Science

CS6200 Information Retrieval Jesse Anderton College of Computer and Information Science

CS6200 Information Retrieval Jesse Anderton College of Computer and Information Science Northeastern University Query Process Review: Ranking Ranking is the process of selecting which documents to show the user, and in what order

866 views • 71 slides
Information Retrieval > Query Us User er Query Words Query Words Search Personalization

Information Retrieval > Query Us User er Query Words Query Words Search Personalization

Information Retrieval > Query Us User er Query Words Query Words Search Personalization Cont ntex ext Ranked List Ranked List Domain Dom Jaime Teevan Cont ntex ext Microsoft Research Ta Task/ sk/Use Use Cont ntex ext

250 views • 7 slides
Certainty Factor certainty factor CF (is the certainty factor in the hypothesis H due to

Certainty Factor certainty factor CF (is the certainty factor in the hypothesis H due to

Certainty Factor certainty factor CF (is the certainty factor in the hypothesis H due to evidence E) ranges between -1 (denial of the hypothesis H) and +1 (confirmation of H) allows the ranking of hypotheses difference between

280 views • 5 slides
user-driven university* ranking (* includes all higher educa(on

user-driven university* ranking (* includes all higher educa(on

The worlds first global, mul(-dimensional, user-driven university* ranking (* includes all higher educa(on ins(tu(ons) www.umul(rank.org Presented by:

467 views • 25 slides
Zero-query information retrieval system no explicit query from user IR triggered by

Zero-query information retrieval system no explicit query from user IR triggered by

Zero-query information retrieval system no explicit query from user IR triggered by context cheap information for user 7 Zero-query information retrieval system 8 Context: Activity 9 Context: Activity Watching football

256 views • 22 slides
Outline of Presentation MARS: Applying Multiplicative Introduction -- the vector model over

Outline of Presentation MARS: Applying Multiplicative Introduction -- the vector model over

Outline of Presentation MARS: Applying Multiplicative Introduction -- the vector model over R+ Adaptive User Preference Multiplicative adaptive query expansion algorithm Retrieval to Web Search MARS -- meta-search engine Zhixiang

47 views • 4 slides
Crawling HTML create an user user inverted index query Search show results inverted

Crawling HTML create an user user inverted index query Search show results inverted

1/22/2013 Standard Web Search Engine Architecture store documents, check for duplicates, extract links crawl the web DocIds Crawling HTML create an user user inverted index query Search show results inverted engine index To user

505 views • 5 slides
Parameter passing Looked at lisps conventional parameter passing mechanism: where

Parameter passing Looked at lisps conventional parameter passing mechanism: where

Parameter passing Looked at lisps conventional parameter passing mechanism: where parameters passed in the order listed, by value, and all parameters are mandatory Several alternatives available: Optional parameters with default

184 views • 14 slides
Visual Recognition and Search April 18, 2008 Joo Hyun Kim Introduction Suppose a stranger in

Visual Recognition and Search April 18, 2008 Joo Hyun Kim Introduction Suppose a stranger in

Visual Recognition and Search April 18, 2008 Joo Hyun Kim Introduction Suppose a stranger in downtown with a tour guide book ?? Austin, TX 2008 04 18 Place Recognition and Kidnapped Robots 2 Introduction Look at guide Whats this?

967 views • 83 slides
Visualization & Visual Analytics 1 Angus Forbes creativecoding.evl.uic.edu/courses/cs424

Visualization & Visual Analytics 1 Angus Forbes creativecoding.evl.uic.edu/courses/cs424

Visualization & Visual Analytics 1 Angus Forbes creativecoding.evl.uic.edu/courses/cs424 How What - Why Why (why is it important or interesting?) What (what visualization tasks will you support?) How (how do the visual encodings /

685 views • 33 slides
CS420 Machine Learning Weinan Zhang Shanghai Jiao Tong University http://wnzhang.net Self

CS420 Machine Learning Weinan Zhang Shanghai Jiao Tong University http://wnzhang.net Self

CS420 Machine Learning Weinan Zhang Shanghai Jiao Tong University http://wnzhang.net Self Introduction Weinan Zhang Position Assistant Professor at CS Dept. of SJTU 2016-now Apex Data and Knowledge Management Lab Research on

1.07k views • 73 slides
LanguageIndependent LanguageIndependent AnswerPrediction AnswerPrediction

LanguageIndependent LanguageIndependent AnswerPrediction AnswerPrediction

LanguageIndependent LanguageIndependent AnswerPrediction AnswerPrediction fromtheWeb fromtheWeb AlejandroFigueroa&G nterNeumann nterNeumann AlejandroFigueroa&G

1.01k views • 22 slides
Br Brand/ nd/ Launc unch h Your ur Ca Career Us Using Li LinkedIn An Anna Centrella

Br Brand/ nd/ Launc unch h Your ur Ca Career Us Using Li LinkedIn An Anna Centrella

Br Brand/ nd/ Launc unch h Your ur Ca Career Us Using Li LinkedIn An Anna Centrella Thayer Profile Sections- How to Start Tips and Hints- Top 5 Profile Building Blocks Ag Agenda The Recruiter View- How to get Found Just Start-

437 views • 29 slides
Homework Watch the SearchEngineLand video: What is SEO

Homework Watch the SearchEngineLand video: What is SEO

Homework Watch the SearchEngineLand video: What is SEO (http://searchengineland.com/guide/what-is-seo) Watch Matt Cutts video: How Search Engines Work (https://www.youtube.com/watch?v=BNHR6IQJGZs) Search Engine Optimization (SEO)

728 views • 24 slides
Back to the sketch-board: Integrating keyword search, semantics, and information retrieval Joel

Back to the sketch-board: Integrating keyword search, semantics, and information retrieval Joel

Back to the sketch-board: Integrating keyword search, semantics, and information retrieval Joel Azzopardi 1 , Fabio Benedetti 2 , Francesco Guerra 2 , and Mihai Lupu 3 1 University of Malta joel.azzopardi@um.edu.mt 2 Universita di Modena e Reggio

450 views • 16 slides

More recommend