SlideShare une entreprise Scribd logo
1  sur  27
Образец заголовка
Tutorial on Query Auto-
Completion
Yichen Feng
feng36 AT illinois DOT edu
University of Illinois at Urbana-
Champaign
Prepared as an assignment for CS410: Text Information Systems in Spring 2016
Образец заголовкаQuery Auto-Completoion
• What is Query Auto-Completion (QAC)
– Giving search suggestions based on typed
prefixes by considering the search history log,
search queries popularity, temporal factors
and personal interests.
Образец заголовкаQAC is important
• Faster users’ input, improve efficiency
• Suggesting possible queries
• Correct users’ typing errors
• Users may not know how to describe the
information he needed
• Speed and Accuracy
• Minimize users’ cognitive and physical
effort
Образец заголовкаQAC is Everywhere
PIAZZA Facebook
Gmail Amazon
USA Government Coursera
Образец заголовкаMost Popular Completion
• Traditional QAC (Most Popular Completion)
– Query are suggested from the previous query
popularity. (Mawarkar and Malemath, 2015)
– Ranked by queries’ number of frequent
occurances
– Data Structure: TRIE
– 𝑀𝐶𝑃 𝒫 = arg max
𝑞∈∁(𝒫)
𝑤 𝑞 , 𝑤 𝑞 =
𝑓(𝑞)
𝑖∈𝒬 𝑓(𝑖)
– Ranked by queries’ number of frequent occurances
– Data Structure: TRIE
– Always treated as baseline
Образец заголовкаQAC Challenges
• Cannot catch the popular temporal topics
• Cannot treat different people differently
• Cannot interact with users’ behaviors (e.g.
clicks)
• Bad performance on the mobile devices
• Needed to be optimized
Образец заголовкаSolutions
• Time-sensitive QAC
– Robust vs. Recent
• Personalized QAC
– User behaviors
– Context based QAC
• Time-sensitive Personalized QAC (Hybrid
model)
• Optimizing search results presentation
• Term by term QAC for mobile search
• QAC for rare prefixes
Образец заголовкаTime-Sensitive QAC
(SIGIR 12)
• Time-sensitive: query popularity changing over time
– “di-”: Dictionary for weekday, Disney for weekend
• Key idea:
– Predicting query popularity
• Forecast quality
• Success & failure analysis
• Temporal model selector
– Rely on shorter but frequent aggregation of data, model
the overall query trends by time-series.
• Method: Time-sensitive auto-completion
– 𝑇𝑆 𝒫, 𝑡 = arg max
𝑞∈∁(𝒫)
𝑤 𝑞|𝑡 , 𝑤 𝑞|𝑡 =
𝑦𝑡(𝑞)
𝑖∈𝒬 𝑦𝑡(𝑖)
– 𝑦𝑡(𝑞): estimated frequency of query q at time t
M. Shokouhi and K. Radinsky. Time-sensitive query auto-completion. In SIGIR ’12, pages 601–610, 2012.
Образец заголовкаTS QAC – Recent vs. Robust
(WWW 14)
• QAC need to sufficiently rank both consistently and recently
popular queries
• Motivation: Finding optimal trade-off between recency and
robustness to achieve better QAC
• Key idea:
– Optimal tradeoff could be researched
– Each query log scenario has different temporal characteristics
• Approaches:
– Based on past popularity distributions
• Maximum Likelihood Estimation, Recent Maximum Likelihood Estimation,
Last N Query Distribution
– Based on short-range predicted query popularity
• Predicted Next N Query Distribution
– Meta approach – optimize the parameters of above apporaches
• Online Parameter Learning
S. Whiting, J. McMinn, and J. Jose. Exploring real-time temporal query auto-completion. In DIR Workshop ’13, pages 12–15
Образец заголовкаPersonalized QAC
(SIGIR 13)
• QAC need to suggest people differently by considering their
own interestes
• Motivation: Queries likelihoods vary drastically between
different demographic groups [Weber and Castillo, 2010] and
individuals [Teevan et al., 2011]
• Key idea:
– Features based on: Users age, gender, location, short- and long-
term history
– Novel supervised framework for leaning to personalize QAC
• Method:
– Similar labelling strategy
• Evaluating by using Mean-Reciprocal-Rank (MRR)
– Learning to rank
• Lambda-MART algorithm (boosted decision trees)
• Location is more effective
M. Shokouhi. Learning to personalize query auto-completion. In SIGIR’13 2013
Образец заголовка
Personalized QAC – Context
Based
(IJARCET 2015)
• Query auto-completer try to accurately predicted what user is typing
• Objective: Improve search quality by predicting the user’s query
based on context
• Key idea:
– Context
• Query similarity
• User’s recent click throughs
• Current location and time
• Keywords and sessions
• Method:
– Most Popular Completion
• Works well when context is empty
– Nearest Completion
• Works well when context exists, terrible when context is empty
– Hybrid Completion
• Combine both MPC and NC
V. Mawarkar and V. Malemath. Context Based Query Auto-Completion. In IJARCET, Volume 4 Issue 6, June 2015.
Образец заголовкаContext Based HCA
(IJARCET 2015)
V. Mawarkar and V. Malemath. Context Based Query Auto-Completion. In IJARCET, Volume 4 Issue 6, June 2015.
Образец заголовкаPersonalized QAC – User Behaviors
(SIGIR14)
• Objective: Explaining the users’ interaction
data to future improving the QAC
performance
• Contributions:
– First set High-resolution QAC query log:
• Recording every keystroke- Enable further analysis on
understanding
– Horizontal skipping bias
• First introduce and unique to QAC
– Vertical position bias
– Two-dimensional Click Model
• Model users’ behavior on PC and mobile devices
Y. Li, A. Dong, H. Wang, H. Deng, Y. Chang, C. Zhai. A Two-dimensional Click Model for Query Auto-completion. In SIGIR’ 2014
Образец заголовкаTwo-Dimensional Click Model
(SIGIR14)
H Model
D Model
Y. Li, A. Dong, H. Wang, H. Deng, Y. Chang, C. Zhai. A Two-dimensional Click Model for Query Auto-completion. In SIGIR’ 2014
Образец заголовкаTime–Sensitive Personalized QAC
(CIKM14)
• Key idea:
– Hybrid model
• Time-sensitivity
• Personalization
– Optimal time window
• Achieving better predition
• Contributions:
– Novel Hybrid Model
– New query popularity prediction method
• Ranking with Mean Reciprocal Rank (MRR)
– Effectiveness analysis
• Significantly outperforms state-of-art time-sensitive
QAC
F. Cai, S. Liang, M. D. Rijke. Time-sensitive Personalized Query Auto-completion. In CIKM’ 2014
Образец заголовкаTSP QAC Performances
(CIKM14)
• Tradeoff between recent and periodicity
– Have critical parameter setting for accuracy
• Baselines check
– Marginally outperforms baselines
• Fact not strongly differential features
– Effective with a longer prefix
– Available evidence matters
• Better QAC ranking
– Sufficient personal queries
– Time-sensitive popularity
F. Cai, S. Liang, M. D. Rijke. Time-sensitive Personalized Query Auto-completion. In CIKM’ 2014
Образец заголовка
Presenting Optimized Search
Results
(WSDM16)
• Objective:
– Selectively presenting query based on a
probabilistic model to achieve optimized search
results presentation
• Key ideas:
– Time-consuming on too many query suggestions
– Measuring the users’ time-loss
– Patient users get more benefits
• Challenges:
– Uncertain factors (e. g. intent, query suggestion
click probabilities)
– Unclear of how long users spend on scanning
M. P. Kato, K. Tanaka. To Suggest, or Not to Suggest for Queries with Diverse Intents: Optimizing Search Result Presentation. In WSDM’ 2016
Образец заголовка
Presenting Optimized Search
Results
(WSDM16)
• Contributions:
– Searcher model
• Interacting with query suggestions
• According to users’ multiple intents
– Optimizing Search Results Presentation (OSRP)
• Mainly focusing on ambiguous or underspecified query
– Examined effects of query suggestion on search
behaviors
• Conducting user survey
– Effectiveness of OSRP
• Patient users
• Queries with limited number of intents
M. P. Kato, K. Tanaka. To Suggest, or Not to Suggest for Queries with Diverse Intents: Optimizing Search Result Presentation. In WSDM’ 2016
Образец заголовкаUsers Survey
(WSDM16)
M. P. Kato, K. Tanaka. To Suggest, or Not to Suggest for Queries with Diverse Intents: Optimizing Search Result Presentation. In WSDM’ 2016
SERP (M. P. Kato and K. Tanaka)
Образец заголовка
Term-by-Term QAC for Mobile
Search
(WSDM16)
• Objective:
– Specialized QAC for mobile search
• Mobile Input:
– Small screen Term-by-Term QAC
– Slower input High quality QAC
– Clumsier QAC matters more than PC
• Key idea:
– Faster exploration of suggestions
– Fits for the text editing in mobile devices
S. Vargas, R. Blanco, P. Mika. Term-by-Term Query Auto-Completion for Mobile Search. In WSDM 2016
Образец заголовкаQuery-Term Graph
(WSDM16)
– Based on previous submitted queries
– Efficient way of
• Storing
• Retrieving
S. Vargas, R. Blanco, P. Mika. Term-by-Term Query Auto-Completion for Mobile Search. In WSDM 2016
Образец заголовкаQAC for Rare Prefixes
(CIKM15)
• Motivation: QAC fail when the prefix is
sufficiently rare
• Key ideas:
– Supervised model ranking synthetic
suggestions
– Query generated by mining query suffixes
– Exploring new ranking signals
• Query n-gram statistics
• Deep convolutional latent semantic model (CLSM)
S. Vargas, R. Blanco, P. Mika. Term-by-Term Query Auto-Completion for Mobile Search. In WSDM 2016
Образец заголовкаModel and Features
(CIKM15)
• LambdaMART model:
– Ranking using features
• N-gram based features
– Model the likelihood that candidate
suggestion is generated by the same LM as
the queries in the search logs
• CLSM based features
– Based on clickthrough data
– Effective for modelling query-document
relevance
– Training on a prefix-suffix pairs datasetB. Mitra, N. Craswell. Query Auto-Completion for Rare Prefixes. In CIKM 2015
Образец заголовкаQAC for Rare Prefixes
(CIKM15)
• Motivation: QAC fail when the prefix is
sufficiently rare
• Key ideas:
– Supervised model ranking synthetic
suggestions
– Query generated by mining query suffixes
– Exploring new ranking signals
• Query n-gram statistics
• Deep convolutional latent semantic model (CLSM)
B. Mitra, N. Craswell. Query Auto-Completion for Rare Prefixes. In CIKM 2015
Образец заголовкаFuture works
• Short range query popularity prediction
• Complex relationships between users’
behavior at different keystrokes
• More complex click models
• Model personalized temporal patterns for
active users (e.g. Professional searchers)
• Online user behavior study on mobile
• Other LM on rare prefixes
Образец заголовкаQAC Development Summary
Образец заголовкаReferences
1. M. Shokouhi and K. Radinsky. Time-sensitive query auto-completion. In SIGIR ’12, pages
601–610, 2012.
2. S. Whiting, J. McMinn, and J. Jose. Exploring real-time temporal query auto-completion. In
DIR Workshop ’13, pages 12–15
3. M. Shokouhi. Learning to personalize query auto-completion. In SIGIR’13 2013
4. V. Mawarkar and V. Malemath. Context Based Query Auto-Completion. In IJARCET, Volume
4 Issue 6, June 2015.
5. Y. Li, A. Dong, H. Wang, H. Deng, Y. Chang, C. Zhai. A Two-dimensional Click Model for
Query Auto-completion. In SIGIR’ 2014
6. F. Cai, S. Liang, M. D. Rijke. Time-sensitive Personalized Query Auto-completion. In CIKM’
2014
7. M. P. Kato, K. Tanaka. To Suggest, or Not to Suggest for Queries with Diverse Intents:
Optimizing Search Result Presentation. In WSDM’ 2016
8. S. Vargas, R. Blanco, P. Mika. Term-by-Term Query Auto-Completion for Mobile Search. In
WSDM 2016
9. B. Mitra, N. Craswell. Query Auto-Completion for Rare Prefixes. In CIKM 2015
10. L. Li, H. Deng, A. Dong, Y. Chang, H. Zha, R. Baeza-Yates. Analyzing User’s Sequential
Behavior in Query Auto-Completion via Markov Processes. In Proc. SIGIR’15 2015.
11. M. Shokouhi. Detecting seasonal queries by time-series analysis. In Proc. SIGIR, pages
1171–1172, Beijing, China, 2011
12. R. W. White and G. Marchionini. Examining the effectiveness of real-time query expansion.
Inf. Process. Manage., 43:685–704, May 2007
13. Z. Bar-Yossef and N. Kraus. Context-sensitive query auto-completion. In WWW ’11, pages
107–116, 2011.

Contenu connexe

Tendances

[AKC2021] agile한 조직으로의 변화 커뮤니티활용사례 (김상기)
[AKC2021] agile한 조직으로의 변화 커뮤니티활용사례 (김상기)[AKC2021] agile한 조직으로의 변화 커뮤니티활용사례 (김상기)
[AKC2021] agile한 조직으로의 변화 커뮤니티활용사례 (김상기)AgileKoreaConference Alliance
 
제10회 보아즈(BOAZ) 빅데이터 컨퍼런스 - 밑바닥부터 시작하는 trivago 추천시스템
제10회 보아즈(BOAZ) 빅데이터 컨퍼런스 - 밑바닥부터 시작하는 trivago 추천시스템제10회 보아즈(BOAZ) 빅데이터 컨퍼런스 - 밑바닥부터 시작하는 trivago 추천시스템
제10회 보아즈(BOAZ) 빅데이터 컨퍼런스 - 밑바닥부터 시작하는 trivago 추천시스템BOAZ Bigdata
 
Search summit-2018-ltr-presentation
Search summit-2018-ltr-presentationSearch summit-2018-ltr-presentation
Search summit-2018-ltr-presentationSujit Pal
 
Recent Trends in Personalization at Netflix
Recent Trends in Personalization at NetflixRecent Trends in Personalization at Netflix
Recent Trends in Personalization at NetflixJustin Basilico
 
Deep Learning for Semantic Search in E-commerce​
Deep Learning for Semantic Search in E-commerce​Deep Learning for Semantic Search in E-commerce​
Deep Learning for Semantic Search in E-commerce​Somnath Banerjee
 
Query Understanding: A Manifesto
Query Understanding: A ManifestoQuery Understanding: A Manifesto
Query Understanding: A ManifestoDaniel Tunkelang
 
Data Analytics Business Intelligence
Data Analytics Business IntelligenceData Analytics Business Intelligence
Data Analytics Business IntelligenceRavikanth-BA
 
La Duck Conf - Initier un Datalab : rien à voir avec ce que j'imaginais !
La Duck Conf - Initier un Datalab : rien à voir avec ce que j'imaginais !La Duck Conf - Initier un Datalab : rien à voir avec ce que j'imaginais !
La Duck Conf - Initier un Datalab : rien à voir avec ce que j'imaginais !OCTO Technology
 
Building a Knowledge Graph using NLP and Ontologies
Building a Knowledge Graph using NLP and OntologiesBuilding a Knowledge Graph using NLP and Ontologies
Building a Knowledge Graph using NLP and OntologiesNeo4j
 
BIG DATA ANALYTICS USING R
BIG DATA ANALYTICS USING  RBIG DATA ANALYTICS USING  R
BIG DATA ANALYTICS USING RUmair Shafique
 
Query Understanding at LinkedIn [Talk at Facebook]
Query Understanding at LinkedIn [Talk at Facebook]Query Understanding at LinkedIn [Talk at Facebook]
Query Understanding at LinkedIn [Talk at Facebook]Abhimanyu Lad
 
기계독해를 위한 BERT 언어처리 모델 활용
기계독해를 위한 BERT 언어처리 모델 활용기계독해를 위한 BERT 언어처리 모델 활용
기계독해를 위한 BERT 언어처리 모델 활용Kenneth Jung
 
Advanced Voice Conversion
Advanced Voice ConversionAdvanced Voice Conversion
Advanced Voice ConversionNU_I_TODALAB
 
Fast, Lenient, and Accurate – Building Personalized Instant Search Experience...
Fast, Lenient, and Accurate – Building Personalized Instant Search Experience...Fast, Lenient, and Accurate – Building Personalized Instant Search Experience...
Fast, Lenient, and Accurate – Building Personalized Instant Search Experience...Abhimanyu Lad
 
한글 검색 질의어 오타 패턴 분석과 사용자 로그를 이용한 질의어 오타 교정 시스템 구축
한글 검색 질의어 오타 패턴 분석과 사용자 로그를 이용한  질의어 오타 교정 시스템 구축한글 검색 질의어 오타 패턴 분석과 사용자 로그를 이용한  질의어 오타 교정 시스템 구축
한글 검색 질의어 오타 패턴 분석과 사용자 로그를 이용한 질의어 오타 교정 시스템 구축Heewon Jeon
 
Information retrieval dynamic indexing
Information retrieval dynamic indexingInformation retrieval dynamic indexing
Information retrieval dynamic indexingNadia Nahar
 

Tendances (20)

[AKC2021] agile한 조직으로의 변화 커뮤니티활용사례 (김상기)
[AKC2021] agile한 조직으로의 변화 커뮤니티활용사례 (김상기)[AKC2021] agile한 조직으로의 변화 커뮤니티활용사례 (김상기)
[AKC2021] agile한 조직으로의 변화 커뮤니티활용사례 (김상기)
 
제10회 보아즈(BOAZ) 빅데이터 컨퍼런스 - 밑바닥부터 시작하는 trivago 추천시스템
제10회 보아즈(BOAZ) 빅데이터 컨퍼런스 - 밑바닥부터 시작하는 trivago 추천시스템제10회 보아즈(BOAZ) 빅데이터 컨퍼런스 - 밑바닥부터 시작하는 trivago 추천시스템
제10회 보아즈(BOAZ) 빅데이터 컨퍼런스 - 밑바닥부터 시작하는 trivago 추천시스템
 
Search summit-2018-ltr-presentation
Search summit-2018-ltr-presentationSearch summit-2018-ltr-presentation
Search summit-2018-ltr-presentation
 
Recent Trends in Personalization at Netflix
Recent Trends in Personalization at NetflixRecent Trends in Personalization at Netflix
Recent Trends in Personalization at Netflix
 
Deep Learning for Semantic Search in E-commerce​
Deep Learning for Semantic Search in E-commerce​Deep Learning for Semantic Search in E-commerce​
Deep Learning for Semantic Search in E-commerce​
 
Query Understanding: A Manifesto
Query Understanding: A ManifestoQuery Understanding: A Manifesto
Query Understanding: A Manifesto
 
Data Analytics Business Intelligence
Data Analytics Business IntelligenceData Analytics Business Intelligence
Data Analytics Business Intelligence
 
La Duck Conf - Initier un Datalab : rien à voir avec ce que j'imaginais !
La Duck Conf - Initier un Datalab : rien à voir avec ce que j'imaginais !La Duck Conf - Initier un Datalab : rien à voir avec ce que j'imaginais !
La Duck Conf - Initier un Datalab : rien à voir avec ce que j'imaginais !
 
Building a Knowledge Graph using NLP and Ontologies
Building a Knowledge Graph using NLP and OntologiesBuilding a Knowledge Graph using NLP and Ontologies
Building a Knowledge Graph using NLP and Ontologies
 
BIG DATA ANALYTICS USING R
BIG DATA ANALYTICS USING  RBIG DATA ANALYTICS USING  R
BIG DATA ANALYTICS USING R
 
Social Listening Tools
Social Listening ToolsSocial Listening Tools
Social Listening Tools
 
Query Understanding at LinkedIn [Talk at Facebook]
Query Understanding at LinkedIn [Talk at Facebook]Query Understanding at LinkedIn [Talk at Facebook]
Query Understanding at LinkedIn [Talk at Facebook]
 
Customer Segmentation
Customer SegmentationCustomer Segmentation
Customer Segmentation
 
기계독해를 위한 BERT 언어처리 모델 활용
기계독해를 위한 BERT 언어처리 모델 활용기계독해를 위한 BERT 언어처리 모델 활용
기계독해를 위한 BERT 언어처리 모델 활용
 
Advanced Voice Conversion
Advanced Voice ConversionAdvanced Voice Conversion
Advanced Voice Conversion
 
Fast, Lenient, and Accurate – Building Personalized Instant Search Experience...
Fast, Lenient, and Accurate – Building Personalized Instant Search Experience...Fast, Lenient, and Accurate – Building Personalized Instant Search Experience...
Fast, Lenient, and Accurate – Building Personalized Instant Search Experience...
 
Information Extraction
Information ExtractionInformation Extraction
Information Extraction
 
한글 검색 질의어 오타 패턴 분석과 사용자 로그를 이용한 질의어 오타 교정 시스템 구축
한글 검색 질의어 오타 패턴 분석과 사용자 로그를 이용한  질의어 오타 교정 시스템 구축한글 검색 질의어 오타 패턴 분석과 사용자 로그를 이용한  질의어 오타 교정 시스템 구축
한글 검색 질의어 오타 패턴 분석과 사용자 로그를 이용한 질의어 오타 교정 시스템 구축
 
Information retrieval dynamic indexing
Information retrieval dynamic indexingInformation retrieval dynamic indexing
Information retrieval dynamic indexing
 
Word2Vec
Word2VecWord2Vec
Word2Vec
 

En vedette

Lucene BootCamp
Lucene BootCampLucene BootCamp
Lucene BootCampGokulD
 
WSDM2016報告会−論文紹介(Beyond Ranking:Optimizing Whole-Page Presentation)#yjwsdm
WSDM2016報告会−論文紹介(Beyond Ranking:Optimizing Whole-Page Presentation)#yjwsdmWSDM2016報告会−論文紹介(Beyond Ranking:Optimizing Whole-Page Presentation)#yjwsdm
WSDM2016報告会−論文紹介(Beyond Ranking:Optimizing Whole-Page Presentation)#yjwsdmYahoo!デベロッパーネットワーク
 
Learning by example: training users through high-quality query suggestions
Learning by example: training users through high-quality query suggestionsLearning by example: training users through high-quality query suggestions
Learning by example: training users through high-quality query suggestionsClaudia Hauff
 
Recurrent networks and beyond by Tomas Mikolov
Recurrent networks and beyond by Tomas MikolovRecurrent networks and beyond by Tomas Mikolov
Recurrent networks and beyond by Tomas MikolovBhaskar Mitra
 
Interleaving - SIGIR 2016 presentation
Interleaving - SIGIR 2016 presentationInterleaving - SIGIR 2016 presentation
Interleaving - SIGIR 2016 presentationXin QIAN
 
Learning to Rank Personalized Search Results in Professional Networks
Learning to Rank Personalized Search Results in Professional NetworksLearning to Rank Personalized Search Results in Professional Networks
Learning to Rank Personalized Search Results in Professional NetworksViet Ha-Thuc
 
Apache Lucene: Searching the Web and Everything Else (Jazoon07)
Apache Lucene: Searching the Web and Everything Else (Jazoon07)Apache Lucene: Searching the Web and Everything Else (Jazoon07)
Apache Lucene: Searching the Web and Everything Else (Jazoon07)dnaber
 
Search Engines Presentation
Search Engines PresentationSearch Engines Presentation
Search Engines PresentationJSCHO9
 
Introduction to Search Engines
Introduction to Search EnginesIntroduction to Search Engines
Introduction to Search EnginesNitin Pande
 
Search Engine Powerpoint
Search Engine PowerpointSearch Engine Powerpoint
Search Engine Powerpoint201014161
 
Recent and Robust Query Auto-Completion - WWW 2014 Conference Presentation
Recent and Robust Query Auto-Completion - WWW 2014 Conference PresentationRecent and Robust Query Auto-Completion - WWW 2014 Conference Presentation
Recent and Robust Query Auto-Completion - WWW 2014 Conference Presentationstewhir
 

En vedette (14)

Lucene BootCamp
Lucene BootCampLucene BootCamp
Lucene BootCamp
 
WSDM2016報告会−論文紹介(Beyond Ranking:Optimizing Whole-Page Presentation)#yjwsdm
WSDM2016報告会−論文紹介(Beyond Ranking:Optimizing Whole-Page Presentation)#yjwsdmWSDM2016報告会−論文紹介(Beyond Ranking:Optimizing Whole-Page Presentation)#yjwsdm
WSDM2016報告会−論文紹介(Beyond Ranking:Optimizing Whole-Page Presentation)#yjwsdm
 
Learning by example: training users through high-quality query suggestions
Learning by example: training users through high-quality query suggestionsLearning by example: training users through high-quality query suggestions
Learning by example: training users through high-quality query suggestions
 
Recurrent networks and beyond by Tomas Mikolov
Recurrent networks and beyond by Tomas MikolovRecurrent networks and beyond by Tomas Mikolov
Recurrent networks and beyond by Tomas Mikolov
 
Interleaving - SIGIR 2016 presentation
Interleaving - SIGIR 2016 presentationInterleaving - SIGIR 2016 presentation
Interleaving - SIGIR 2016 presentation
 
Learning to Rank Personalized Search Results in Professional Networks
Learning to Rank Personalized Search Results in Professional NetworksLearning to Rank Personalized Search Results in Professional Networks
Learning to Rank Personalized Search Results in Professional Networks
 
Lucene basics
Lucene basicsLucene basics
Lucene basics
 
Apache Lucene: Searching the Web and Everything Else (Jazoon07)
Apache Lucene: Searching the Web and Everything Else (Jazoon07)Apache Lucene: Searching the Web and Everything Else (Jazoon07)
Apache Lucene: Searching the Web and Everything Else (Jazoon07)
 
Search engine
Search engineSearch engine
Search engine
 
Search Engines Presentation
Search Engines PresentationSearch Engines Presentation
Search Engines Presentation
 
Introduction to Search Engines
Introduction to Search EnginesIntroduction to Search Engines
Introduction to Search Engines
 
Search engines
Search enginesSearch engines
Search engines
 
Search Engine Powerpoint
Search Engine PowerpointSearch Engine Powerpoint
Search Engine Powerpoint
 
Recent and Robust Query Auto-Completion - WWW 2014 Conference Presentation
Recent and Robust Query Auto-Completion - WWW 2014 Conference PresentationRecent and Robust Query Auto-Completion - WWW 2014 Conference Presentation
Recent and Robust Query Auto-Completion - WWW 2014 Conference Presentation
 

Similaire à Tutorial on query auto-completion

Semantic Need: Guiding Metadata Annotations by Questions People #ask
Semantic Need: Guiding Metadata Annotations by Questions People #askSemantic Need: Guiding Metadata Annotations by Questions People #ask
Semantic Need: Guiding Metadata Annotations by Questions People #askHans-Joerg Happel
 
Multimedia Answer Generation for Community Question Answering
Multimedia Answer Generation for Community Question AnsweringMultimedia Answer Generation for Community Question Answering
Multimedia Answer Generation for Community Question AnsweringSWAMI06
 
Exploring Session Context using Distributed Representations of Queries and Re...
Exploring Session Context using Distributed Representations of Queries and Re...Exploring Session Context using Distributed Representations of Queries and Re...
Exploring Session Context using Distributed Representations of Queries and Re...Bhaskar Mitra
 
Top-k Exploration of Query Candidates for Efficient Keyword Search on Graph-S...
Top-k Exploration of Query Candidates for Efficient Keyword Search on Graph-S...Top-k Exploration of Query Candidates for Efficient Keyword Search on Graph-S...
Top-k Exploration of Query Candidates for Efficient Keyword Search on Graph-S...Thanh Tran
 
Beyond Collaborative Filtering: Learning to Rank Research Articles
Beyond Collaborative Filtering: Learning to Rank Research ArticlesBeyond Collaborative Filtering: Learning to Rank Research Articles
Beyond Collaborative Filtering: Learning to Rank Research ArticlesMaya Hristakeva
 
Taking the Pain out of Data Science - RecSys Machine Learning Framework Over ...
Taking the Pain out of Data Science - RecSys Machine Learning Framework Over ...Taking the Pain out of Data Science - RecSys Machine Learning Framework Over ...
Taking the Pain out of Data Science - RecSys Machine Learning Framework Over ...Sonya Liberman
 
Best Practices for Large Scale Text Mining Processing
Best Practices for Large Scale Text Mining ProcessingBest Practices for Large Scale Text Mining Processing
Best Practices for Large Scale Text Mining ProcessingOntotext
 
Tag based recommender system
Tag based recommender systemTag based recommender system
Tag based recommender systemKaren Li
 
A Machine Learning Approach to SPARQL Query Performance Prediction
A Machine Learning Approach to SPARQL Query Performance PredictionA Machine Learning Approach to SPARQL Query Performance Prediction
A Machine Learning Approach to SPARQL Query Performance PredictionRakebul Hasan
 
Leveraging Dynamic Query Subtopics for Time-aware Search Result Diversification
Leveraging Dynamic Query Subtopics for Time-aware Search Result DiversificationLeveraging Dynamic Query Subtopics for Time-aware Search Result Diversification
Leveraging Dynamic Query Subtopics for Time-aware Search Result DiversificationNattiya Kanhabua
 
Utilizing Marginal Net Utility for Recommendation in E-commerce
Utilizing Marginal Net Utility for Recommendation in E-commerceUtilizing Marginal Net Utility for Recommendation in E-commerce
Utilizing Marginal Net Utility for Recommendation in E-commerceLiangjie Hong
 
The Analytics Frontier of the Hadoop Eco-System
The Analytics Frontier of the Hadoop Eco-SystemThe Analytics Frontier of the Hadoop Eco-System
The Analytics Frontier of the Hadoop Eco-Systeminside-BigData.com
 
Tangram: Distributed Scheduling Framework for Apache Spark at Facebook
Tangram: Distributed Scheduling Framework for Apache Spark at FacebookTangram: Distributed Scheduling Framework for Apache Spark at Facebook
Tangram: Distributed Scheduling Framework for Apache Spark at FacebookDatabricks
 
Relevancy and Search Quality Analysis - Search Technologies
Relevancy and Search Quality Analysis - Search TechnologiesRelevancy and Search Quality Analysis - Search Technologies
Relevancy and Search Quality Analysis - Search Technologiesenterprisesearchmeetup
 
Enhancing Information Retrieval by Personalization Techniques
Enhancing Information Retrieval by Personalization TechniquesEnhancing Information Retrieval by Personalization Techniques
Enhancing Information Retrieval by Personalization Techniquesveningstonk
 
RecSys 2015 Tutorial – Scalable Recommender Systems: Where Machine Learning...
 RecSys 2015 Tutorial – Scalable Recommender Systems: Where Machine Learning... RecSys 2015 Tutorial – Scalable Recommender Systems: Where Machine Learning...
RecSys 2015 Tutorial – Scalable Recommender Systems: Where Machine Learning...S. Diana Hu
 
RecSys 2015 Tutorial - Scalable Recommender Systems: Where Machine Learning m...
RecSys 2015 Tutorial - Scalable Recommender Systems: Where Machine Learning m...RecSys 2015 Tutorial - Scalable Recommender Systems: Where Machine Learning m...
RecSys 2015 Tutorial - Scalable Recommender Systems: Where Machine Learning m...Joaquin Delgado PhD.
 
MUDROD - Mining and Utilizing Dataset Relevancy from Oceanographic Dataset Me...
MUDROD - Mining and Utilizing Dataset Relevancy from Oceanographic Dataset Me...MUDROD - Mining and Utilizing Dataset Relevancy from Oceanographic Dataset Me...
MUDROD - Mining and Utilizing Dataset Relevancy from Oceanographic Dataset Me...Yongyao Jiang
 
Building High Available and Scalable Machine Learning Applications
Building High Available and Scalable Machine Learning ApplicationsBuilding High Available and Scalable Machine Learning Applications
Building High Available and Scalable Machine Learning ApplicationsYalçın Yenigün
 
Clickstream Analytics with Markov Chains
Clickstream Analytics with Markov Chains Clickstream Analytics with Markov Chains
Clickstream Analytics with Markov Chains Alex Papageorgiou
 

Similaire à Tutorial on query auto-completion (20)

Semantic Need: Guiding Metadata Annotations by Questions People #ask
Semantic Need: Guiding Metadata Annotations by Questions People #askSemantic Need: Guiding Metadata Annotations by Questions People #ask
Semantic Need: Guiding Metadata Annotations by Questions People #ask
 
Multimedia Answer Generation for Community Question Answering
Multimedia Answer Generation for Community Question AnsweringMultimedia Answer Generation for Community Question Answering
Multimedia Answer Generation for Community Question Answering
 
Exploring Session Context using Distributed Representations of Queries and Re...
Exploring Session Context using Distributed Representations of Queries and Re...Exploring Session Context using Distributed Representations of Queries and Re...
Exploring Session Context using Distributed Representations of Queries and Re...
 
Top-k Exploration of Query Candidates for Efficient Keyword Search on Graph-S...
Top-k Exploration of Query Candidates for Efficient Keyword Search on Graph-S...Top-k Exploration of Query Candidates for Efficient Keyword Search on Graph-S...
Top-k Exploration of Query Candidates for Efficient Keyword Search on Graph-S...
 
Beyond Collaborative Filtering: Learning to Rank Research Articles
Beyond Collaborative Filtering: Learning to Rank Research ArticlesBeyond Collaborative Filtering: Learning to Rank Research Articles
Beyond Collaborative Filtering: Learning to Rank Research Articles
 
Taking the Pain out of Data Science - RecSys Machine Learning Framework Over ...
Taking the Pain out of Data Science - RecSys Machine Learning Framework Over ...Taking the Pain out of Data Science - RecSys Machine Learning Framework Over ...
Taking the Pain out of Data Science - RecSys Machine Learning Framework Over ...
 
Best Practices for Large Scale Text Mining Processing
Best Practices for Large Scale Text Mining ProcessingBest Practices for Large Scale Text Mining Processing
Best Practices for Large Scale Text Mining Processing
 
Tag based recommender system
Tag based recommender systemTag based recommender system
Tag based recommender system
 
A Machine Learning Approach to SPARQL Query Performance Prediction
A Machine Learning Approach to SPARQL Query Performance PredictionA Machine Learning Approach to SPARQL Query Performance Prediction
A Machine Learning Approach to SPARQL Query Performance Prediction
 
Leveraging Dynamic Query Subtopics for Time-aware Search Result Diversification
Leveraging Dynamic Query Subtopics for Time-aware Search Result DiversificationLeveraging Dynamic Query Subtopics for Time-aware Search Result Diversification
Leveraging Dynamic Query Subtopics for Time-aware Search Result Diversification
 
Utilizing Marginal Net Utility for Recommendation in E-commerce
Utilizing Marginal Net Utility for Recommendation in E-commerceUtilizing Marginal Net Utility for Recommendation in E-commerce
Utilizing Marginal Net Utility for Recommendation in E-commerce
 
The Analytics Frontier of the Hadoop Eco-System
The Analytics Frontier of the Hadoop Eco-SystemThe Analytics Frontier of the Hadoop Eco-System
The Analytics Frontier of the Hadoop Eco-System
 
Tangram: Distributed Scheduling Framework for Apache Spark at Facebook
Tangram: Distributed Scheduling Framework for Apache Spark at FacebookTangram: Distributed Scheduling Framework for Apache Spark at Facebook
Tangram: Distributed Scheduling Framework for Apache Spark at Facebook
 
Relevancy and Search Quality Analysis - Search Technologies
Relevancy and Search Quality Analysis - Search TechnologiesRelevancy and Search Quality Analysis - Search Technologies
Relevancy and Search Quality Analysis - Search Technologies
 
Enhancing Information Retrieval by Personalization Techniques
Enhancing Information Retrieval by Personalization TechniquesEnhancing Information Retrieval by Personalization Techniques
Enhancing Information Retrieval by Personalization Techniques
 
RecSys 2015 Tutorial – Scalable Recommender Systems: Where Machine Learning...
 RecSys 2015 Tutorial – Scalable Recommender Systems: Where Machine Learning... RecSys 2015 Tutorial – Scalable Recommender Systems: Where Machine Learning...
RecSys 2015 Tutorial – Scalable Recommender Systems: Where Machine Learning...
 
RecSys 2015 Tutorial - Scalable Recommender Systems: Where Machine Learning m...
RecSys 2015 Tutorial - Scalable Recommender Systems: Where Machine Learning m...RecSys 2015 Tutorial - Scalable Recommender Systems: Where Machine Learning m...
RecSys 2015 Tutorial - Scalable Recommender Systems: Where Machine Learning m...
 
MUDROD - Mining and Utilizing Dataset Relevancy from Oceanographic Dataset Me...
MUDROD - Mining and Utilizing Dataset Relevancy from Oceanographic Dataset Me...MUDROD - Mining and Utilizing Dataset Relevancy from Oceanographic Dataset Me...
MUDROD - Mining and Utilizing Dataset Relevancy from Oceanographic Dataset Me...
 
Building High Available and Scalable Machine Learning Applications
Building High Available and Scalable Machine Learning ApplicationsBuilding High Available and Scalable Machine Learning Applications
Building High Available and Scalable Machine Learning Applications
 
Clickstream Analytics with Markov Chains
Clickstream Analytics with Markov Chains Clickstream Analytics with Markov Chains
Clickstream Analytics with Markov Chains
 

Dernier

Correctly Loading Incremental Data at Scale
Correctly Loading Incremental Data at ScaleCorrectly Loading Incremental Data at Scale
Correctly Loading Incremental Data at ScaleAlluxio, Inc.
 
Work Experience-Dalton Park.pptxfvvvvvvv
Work Experience-Dalton Park.pptxfvvvvvvvWork Experience-Dalton Park.pptxfvvvvvvv
Work Experience-Dalton Park.pptxfvvvvvvvLewisJB
 
CCS355 Neural Network & Deep Learning UNIT III notes and Question bank .pdf
CCS355 Neural Network & Deep Learning UNIT III notes and Question bank .pdfCCS355 Neural Network & Deep Learning UNIT III notes and Question bank .pdf
CCS355 Neural Network & Deep Learning UNIT III notes and Question bank .pdfAsst.prof M.Gokilavani
 
Transport layer issues and challenges - Guide
Transport layer issues and challenges - GuideTransport layer issues and challenges - Guide
Transport layer issues and challenges - GuideGOPINATHS437943
 
Gurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort service
Gurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort serviceGurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort service
Gurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort servicejennyeacort
 
INFLUENCE OF NANOSILICA ON THE PROPERTIES OF CONCRETE
INFLUENCE OF NANOSILICA ON THE PROPERTIES OF CONCRETEINFLUENCE OF NANOSILICA ON THE PROPERTIES OF CONCRETE
INFLUENCE OF NANOSILICA ON THE PROPERTIES OF CONCRETEroselinkalist12
 
CCS355 Neural Network & Deep Learning Unit II Notes with Question bank .pdf
CCS355 Neural Network & Deep Learning Unit II Notes with Question bank .pdfCCS355 Neural Network & Deep Learning Unit II Notes with Question bank .pdf
CCS355 Neural Network & Deep Learning Unit II Notes with Question bank .pdfAsst.prof M.Gokilavani
 
welding defects observed during the welding
welding defects observed during the weldingwelding defects observed during the welding
welding defects observed during the weldingMuhammadUzairLiaqat
 
Class 1 | NFPA 72 | Overview Fire Alarm System
Class 1 | NFPA 72 | Overview Fire Alarm SystemClass 1 | NFPA 72 | Overview Fire Alarm System
Class 1 | NFPA 72 | Overview Fire Alarm Systemirfanmechengr
 
Piping Basic stress analysis by engineering
Piping Basic stress analysis by engineeringPiping Basic stress analysis by engineering
Piping Basic stress analysis by engineeringJuanCarlosMorales19600
 
8251 universal synchronous asynchronous receiver transmitter
8251 universal synchronous asynchronous receiver transmitter8251 universal synchronous asynchronous receiver transmitter
8251 universal synchronous asynchronous receiver transmitterShivangiSharma879191
 
US Department of Education FAFSA Week of Action
US Department of Education FAFSA Week of ActionUS Department of Education FAFSA Week of Action
US Department of Education FAFSA Week of ActionMebane Rash
 
Arduino_CSE ece ppt for working and principal of arduino.ppt
Arduino_CSE ece ppt for working and principal of arduino.pptArduino_CSE ece ppt for working and principal of arduino.ppt
Arduino_CSE ece ppt for working and principal of arduino.pptSAURABHKUMAR892774
 
Electronically Controlled suspensions system .pdf
Electronically Controlled suspensions system .pdfElectronically Controlled suspensions system .pdf
Electronically Controlled suspensions system .pdfme23b1001
 
An experimental study in using natural admixture as an alternative for chemic...
An experimental study in using natural admixture as an alternative for chemic...An experimental study in using natural admixture as an alternative for chemic...
An experimental study in using natural admixture as an alternative for chemic...Chandu841456
 
Indian Dairy Industry Present Status and.ppt
Indian Dairy Industry Present Status and.pptIndian Dairy Industry Present Status and.ppt
Indian Dairy Industry Present Status and.pptMadan Karki
 
TechTAC® CFD Report Summary: A Comparison of Two Types of Tubing Anchor Catchers
TechTAC® CFD Report Summary: A Comparison of Two Types of Tubing Anchor CatchersTechTAC® CFD Report Summary: A Comparison of Two Types of Tubing Anchor Catchers
TechTAC® CFD Report Summary: A Comparison of Two Types of Tubing Anchor Catcherssdickerson1
 

Dernier (20)

Correctly Loading Incremental Data at Scale
Correctly Loading Incremental Data at ScaleCorrectly Loading Incremental Data at Scale
Correctly Loading Incremental Data at Scale
 
Work Experience-Dalton Park.pptxfvvvvvvv
Work Experience-Dalton Park.pptxfvvvvvvvWork Experience-Dalton Park.pptxfvvvvvvv
Work Experience-Dalton Park.pptxfvvvvvvv
 
CCS355 Neural Network & Deep Learning UNIT III notes and Question bank .pdf
CCS355 Neural Network & Deep Learning UNIT III notes and Question bank .pdfCCS355 Neural Network & Deep Learning UNIT III notes and Question bank .pdf
CCS355 Neural Network & Deep Learning UNIT III notes and Question bank .pdf
 
Transport layer issues and challenges - Guide
Transport layer issues and challenges - GuideTransport layer issues and challenges - Guide
Transport layer issues and challenges - Guide
 
Gurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort service
Gurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort serviceGurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort service
Gurgaon ✡️9711147426✨Call In girls Gurgaon Sector 51 escort service
 
INFLUENCE OF NANOSILICA ON THE PROPERTIES OF CONCRETE
INFLUENCE OF NANOSILICA ON THE PROPERTIES OF CONCRETEINFLUENCE OF NANOSILICA ON THE PROPERTIES OF CONCRETE
INFLUENCE OF NANOSILICA ON THE PROPERTIES OF CONCRETE
 
CCS355 Neural Network & Deep Learning Unit II Notes with Question bank .pdf
CCS355 Neural Network & Deep Learning Unit II Notes with Question bank .pdfCCS355 Neural Network & Deep Learning Unit II Notes with Question bank .pdf
CCS355 Neural Network & Deep Learning Unit II Notes with Question bank .pdf
 
welding defects observed during the welding
welding defects observed during the weldingwelding defects observed during the welding
welding defects observed during the welding
 
Class 1 | NFPA 72 | Overview Fire Alarm System
Class 1 | NFPA 72 | Overview Fire Alarm SystemClass 1 | NFPA 72 | Overview Fire Alarm System
Class 1 | NFPA 72 | Overview Fire Alarm System
 
Piping Basic stress analysis by engineering
Piping Basic stress analysis by engineeringPiping Basic stress analysis by engineering
Piping Basic stress analysis by engineering
 
8251 universal synchronous asynchronous receiver transmitter
8251 universal synchronous asynchronous receiver transmitter8251 universal synchronous asynchronous receiver transmitter
8251 universal synchronous asynchronous receiver transmitter
 
US Department of Education FAFSA Week of Action
US Department of Education FAFSA Week of ActionUS Department of Education FAFSA Week of Action
US Department of Education FAFSA Week of Action
 
Arduino_CSE ece ppt for working and principal of arduino.ppt
Arduino_CSE ece ppt for working and principal of arduino.pptArduino_CSE ece ppt for working and principal of arduino.ppt
Arduino_CSE ece ppt for working and principal of arduino.ppt
 
Exploring_Network_Security_with_JA3_by_Rakesh Seal.pptx
Exploring_Network_Security_with_JA3_by_Rakesh Seal.pptxExploring_Network_Security_with_JA3_by_Rakesh Seal.pptx
Exploring_Network_Security_with_JA3_by_Rakesh Seal.pptx
 
POWER SYSTEMS-1 Complete notes examples
POWER SYSTEMS-1 Complete notes  examplesPOWER SYSTEMS-1 Complete notes  examples
POWER SYSTEMS-1 Complete notes examples
 
Electronically Controlled suspensions system .pdf
Electronically Controlled suspensions system .pdfElectronically Controlled suspensions system .pdf
Electronically Controlled suspensions system .pdf
 
An experimental study in using natural admixture as an alternative for chemic...
An experimental study in using natural admixture as an alternative for chemic...An experimental study in using natural admixture as an alternative for chemic...
An experimental study in using natural admixture as an alternative for chemic...
 
young call girls in Rajiv Chowk🔝 9953056974 🔝 Delhi escort Service
young call girls in Rajiv Chowk🔝 9953056974 🔝 Delhi escort Serviceyoung call girls in Rajiv Chowk🔝 9953056974 🔝 Delhi escort Service
young call girls in Rajiv Chowk🔝 9953056974 🔝 Delhi escort Service
 
Indian Dairy Industry Present Status and.ppt
Indian Dairy Industry Present Status and.pptIndian Dairy Industry Present Status and.ppt
Indian Dairy Industry Present Status and.ppt
 
TechTAC® CFD Report Summary: A Comparison of Two Types of Tubing Anchor Catchers
TechTAC® CFD Report Summary: A Comparison of Two Types of Tubing Anchor CatchersTechTAC® CFD Report Summary: A Comparison of Two Types of Tubing Anchor Catchers
TechTAC® CFD Report Summary: A Comparison of Two Types of Tubing Anchor Catchers
 

Tutorial on query auto-completion

  • 1. Образец заголовка Tutorial on Query Auto- Completion Yichen Feng feng36 AT illinois DOT edu University of Illinois at Urbana- Champaign Prepared as an assignment for CS410: Text Information Systems in Spring 2016
  • 2. Образец заголовкаQuery Auto-Completoion • What is Query Auto-Completion (QAC) – Giving search suggestions based on typed prefixes by considering the search history log, search queries popularity, temporal factors and personal interests.
  • 3. Образец заголовкаQAC is important • Faster users’ input, improve efficiency • Suggesting possible queries • Correct users’ typing errors • Users may not know how to describe the information he needed • Speed and Accuracy • Minimize users’ cognitive and physical effort
  • 4. Образец заголовкаQAC is Everywhere PIAZZA Facebook Gmail Amazon USA Government Coursera
  • 5. Образец заголовкаMost Popular Completion • Traditional QAC (Most Popular Completion) – Query are suggested from the previous query popularity. (Mawarkar and Malemath, 2015) – Ranked by queries’ number of frequent occurances – Data Structure: TRIE – 𝑀𝐶𝑃 𝒫 = arg max 𝑞∈∁(𝒫) 𝑤 𝑞 , 𝑤 𝑞 = 𝑓(𝑞) 𝑖∈𝒬 𝑓(𝑖) – Ranked by queries’ number of frequent occurances – Data Structure: TRIE – Always treated as baseline
  • 6. Образец заголовкаQAC Challenges • Cannot catch the popular temporal topics • Cannot treat different people differently • Cannot interact with users’ behaviors (e.g. clicks) • Bad performance on the mobile devices • Needed to be optimized
  • 7. Образец заголовкаSolutions • Time-sensitive QAC – Robust vs. Recent • Personalized QAC – User behaviors – Context based QAC • Time-sensitive Personalized QAC (Hybrid model) • Optimizing search results presentation • Term by term QAC for mobile search • QAC for rare prefixes
  • 8. Образец заголовкаTime-Sensitive QAC (SIGIR 12) • Time-sensitive: query popularity changing over time – “di-”: Dictionary for weekday, Disney for weekend • Key idea: – Predicting query popularity • Forecast quality • Success & failure analysis • Temporal model selector – Rely on shorter but frequent aggregation of data, model the overall query trends by time-series. • Method: Time-sensitive auto-completion – 𝑇𝑆 𝒫, 𝑡 = arg max 𝑞∈∁(𝒫) 𝑤 𝑞|𝑡 , 𝑤 𝑞|𝑡 = 𝑦𝑡(𝑞) 𝑖∈𝒬 𝑦𝑡(𝑖) – 𝑦𝑡(𝑞): estimated frequency of query q at time t M. Shokouhi and K. Radinsky. Time-sensitive query auto-completion. In SIGIR ’12, pages 601–610, 2012.
  • 9. Образец заголовкаTS QAC – Recent vs. Robust (WWW 14) • QAC need to sufficiently rank both consistently and recently popular queries • Motivation: Finding optimal trade-off between recency and robustness to achieve better QAC • Key idea: – Optimal tradeoff could be researched – Each query log scenario has different temporal characteristics • Approaches: – Based on past popularity distributions • Maximum Likelihood Estimation, Recent Maximum Likelihood Estimation, Last N Query Distribution – Based on short-range predicted query popularity • Predicted Next N Query Distribution – Meta approach – optimize the parameters of above apporaches • Online Parameter Learning S. Whiting, J. McMinn, and J. Jose. Exploring real-time temporal query auto-completion. In DIR Workshop ’13, pages 12–15
  • 10. Образец заголовкаPersonalized QAC (SIGIR 13) • QAC need to suggest people differently by considering their own interestes • Motivation: Queries likelihoods vary drastically between different demographic groups [Weber and Castillo, 2010] and individuals [Teevan et al., 2011] • Key idea: – Features based on: Users age, gender, location, short- and long- term history – Novel supervised framework for leaning to personalize QAC • Method: – Similar labelling strategy • Evaluating by using Mean-Reciprocal-Rank (MRR) – Learning to rank • Lambda-MART algorithm (boosted decision trees) • Location is more effective M. Shokouhi. Learning to personalize query auto-completion. In SIGIR’13 2013
  • 11. Образец заголовка Personalized QAC – Context Based (IJARCET 2015) • Query auto-completer try to accurately predicted what user is typing • Objective: Improve search quality by predicting the user’s query based on context • Key idea: – Context • Query similarity • User’s recent click throughs • Current location and time • Keywords and sessions • Method: – Most Popular Completion • Works well when context is empty – Nearest Completion • Works well when context exists, terrible when context is empty – Hybrid Completion • Combine both MPC and NC V. Mawarkar and V. Malemath. Context Based Query Auto-Completion. In IJARCET, Volume 4 Issue 6, June 2015.
  • 12. Образец заголовкаContext Based HCA (IJARCET 2015) V. Mawarkar and V. Malemath. Context Based Query Auto-Completion. In IJARCET, Volume 4 Issue 6, June 2015.
  • 13. Образец заголовкаPersonalized QAC – User Behaviors (SIGIR14) • Objective: Explaining the users’ interaction data to future improving the QAC performance • Contributions: – First set High-resolution QAC query log: • Recording every keystroke- Enable further analysis on understanding – Horizontal skipping bias • First introduce and unique to QAC – Vertical position bias – Two-dimensional Click Model • Model users’ behavior on PC and mobile devices Y. Li, A. Dong, H. Wang, H. Deng, Y. Chang, C. Zhai. A Two-dimensional Click Model for Query Auto-completion. In SIGIR’ 2014
  • 14. Образец заголовкаTwo-Dimensional Click Model (SIGIR14) H Model D Model Y. Li, A. Dong, H. Wang, H. Deng, Y. Chang, C. Zhai. A Two-dimensional Click Model for Query Auto-completion. In SIGIR’ 2014
  • 15. Образец заголовкаTime–Sensitive Personalized QAC (CIKM14) • Key idea: – Hybrid model • Time-sensitivity • Personalization – Optimal time window • Achieving better predition • Contributions: – Novel Hybrid Model – New query popularity prediction method • Ranking with Mean Reciprocal Rank (MRR) – Effectiveness analysis • Significantly outperforms state-of-art time-sensitive QAC F. Cai, S. Liang, M. D. Rijke. Time-sensitive Personalized Query Auto-completion. In CIKM’ 2014
  • 16. Образец заголовкаTSP QAC Performances (CIKM14) • Tradeoff between recent and periodicity – Have critical parameter setting for accuracy • Baselines check – Marginally outperforms baselines • Fact not strongly differential features – Effective with a longer prefix – Available evidence matters • Better QAC ranking – Sufficient personal queries – Time-sensitive popularity F. Cai, S. Liang, M. D. Rijke. Time-sensitive Personalized Query Auto-completion. In CIKM’ 2014
  • 17. Образец заголовка Presenting Optimized Search Results (WSDM16) • Objective: – Selectively presenting query based on a probabilistic model to achieve optimized search results presentation • Key ideas: – Time-consuming on too many query suggestions – Measuring the users’ time-loss – Patient users get more benefits • Challenges: – Uncertain factors (e. g. intent, query suggestion click probabilities) – Unclear of how long users spend on scanning M. P. Kato, K. Tanaka. To Suggest, or Not to Suggest for Queries with Diverse Intents: Optimizing Search Result Presentation. In WSDM’ 2016
  • 18. Образец заголовка Presenting Optimized Search Results (WSDM16) • Contributions: – Searcher model • Interacting with query suggestions • According to users’ multiple intents – Optimizing Search Results Presentation (OSRP) • Mainly focusing on ambiguous or underspecified query – Examined effects of query suggestion on search behaviors • Conducting user survey – Effectiveness of OSRP • Patient users • Queries with limited number of intents M. P. Kato, K. Tanaka. To Suggest, or Not to Suggest for Queries with Diverse Intents: Optimizing Search Result Presentation. In WSDM’ 2016
  • 19. Образец заголовкаUsers Survey (WSDM16) M. P. Kato, K. Tanaka. To Suggest, or Not to Suggest for Queries with Diverse Intents: Optimizing Search Result Presentation. In WSDM’ 2016 SERP (M. P. Kato and K. Tanaka)
  • 20. Образец заголовка Term-by-Term QAC for Mobile Search (WSDM16) • Objective: – Specialized QAC for mobile search • Mobile Input: – Small screen Term-by-Term QAC – Slower input High quality QAC – Clumsier QAC matters more than PC • Key idea: – Faster exploration of suggestions – Fits for the text editing in mobile devices S. Vargas, R. Blanco, P. Mika. Term-by-Term Query Auto-Completion for Mobile Search. In WSDM 2016
  • 21. Образец заголовкаQuery-Term Graph (WSDM16) – Based on previous submitted queries – Efficient way of • Storing • Retrieving S. Vargas, R. Blanco, P. Mika. Term-by-Term Query Auto-Completion for Mobile Search. In WSDM 2016
  • 22. Образец заголовкаQAC for Rare Prefixes (CIKM15) • Motivation: QAC fail when the prefix is sufficiently rare • Key ideas: – Supervised model ranking synthetic suggestions – Query generated by mining query suffixes – Exploring new ranking signals • Query n-gram statistics • Deep convolutional latent semantic model (CLSM) S. Vargas, R. Blanco, P. Mika. Term-by-Term Query Auto-Completion for Mobile Search. In WSDM 2016
  • 23. Образец заголовкаModel and Features (CIKM15) • LambdaMART model: – Ranking using features • N-gram based features – Model the likelihood that candidate suggestion is generated by the same LM as the queries in the search logs • CLSM based features – Based on clickthrough data – Effective for modelling query-document relevance – Training on a prefix-suffix pairs datasetB. Mitra, N. Craswell. Query Auto-Completion for Rare Prefixes. In CIKM 2015
  • 24. Образец заголовкаQAC for Rare Prefixes (CIKM15) • Motivation: QAC fail when the prefix is sufficiently rare • Key ideas: – Supervised model ranking synthetic suggestions – Query generated by mining query suffixes – Exploring new ranking signals • Query n-gram statistics • Deep convolutional latent semantic model (CLSM) B. Mitra, N. Craswell. Query Auto-Completion for Rare Prefixes. In CIKM 2015
  • 25. Образец заголовкаFuture works • Short range query popularity prediction • Complex relationships between users’ behavior at different keystrokes • More complex click models • Model personalized temporal patterns for active users (e.g. Professional searchers) • Online user behavior study on mobile • Other LM on rare prefixes
  • 27. Образец заголовкаReferences 1. M. Shokouhi and K. Radinsky. Time-sensitive query auto-completion. In SIGIR ’12, pages 601–610, 2012. 2. S. Whiting, J. McMinn, and J. Jose. Exploring real-time temporal query auto-completion. In DIR Workshop ’13, pages 12–15 3. M. Shokouhi. Learning to personalize query auto-completion. In SIGIR’13 2013 4. V. Mawarkar and V. Malemath. Context Based Query Auto-Completion. In IJARCET, Volume 4 Issue 6, June 2015. 5. Y. Li, A. Dong, H. Wang, H. Deng, Y. Chang, C. Zhai. A Two-dimensional Click Model for Query Auto-completion. In SIGIR’ 2014 6. F. Cai, S. Liang, M. D. Rijke. Time-sensitive Personalized Query Auto-completion. In CIKM’ 2014 7. M. P. Kato, K. Tanaka. To Suggest, or Not to Suggest for Queries with Diverse Intents: Optimizing Search Result Presentation. In WSDM’ 2016 8. S. Vargas, R. Blanco, P. Mika. Term-by-Term Query Auto-Completion for Mobile Search. In WSDM 2016 9. B. Mitra, N. Craswell. Query Auto-Completion for Rare Prefixes. In CIKM 2015 10. L. Li, H. Deng, A. Dong, Y. Chang, H. Zha, R. Baeza-Yates. Analyzing User’s Sequential Behavior in Query Auto-Completion via Markov Processes. In Proc. SIGIR’15 2015. 11. M. Shokouhi. Detecting seasonal queries by time-series analysis. In Proc. SIGIR, pages 1171–1172, Beijing, China, 2011 12. R. W. White and G. Marchionini. Examining the effectiveness of real-time query expansion. Inf. Process. Manage., 43:685–704, May 2007 13. Z. Bar-Yossef and N. Kraus. Context-sensitive query auto-completion. In WWW ’11, pages 107–116, 2011.