Personalised Search on the World Wide Web Originally by Micarelli, Gasparetti, Sciarrone & Gauch

Slides:



Advertisements
Similar presentations
Taxonomy & Ontology Impact on Search Infrastructure John R. McGrath Sr. Director, Fast Search & Transfer.
Advertisements

Personalised Search on the World Wide Web Presented by: Team Grape.
A Domain Level Personalization Technique A. Campi, M. Mazuran, S. Ronchi.
Haystack: Per-User Information Environment 1999 Conference on Information and Knowledge Management Eytan Adar et al Presented by Xiao Hu CS491CXZ.
Natural Language Processing WEB SEARCH ENGINES August, 2002.
Web Mining Research: A Survey Authors: Raymond Kosala & Hendrik Blockeel Presenter: Ryan Patterson April 23rd 2014 CS332 Data Mining pg 01.
Information Retrieval: Human-Computer Interfaces and Information Access Process.
Research topics Semantic Web - Spring 2007 Computer Engineering Department Sharif University of Technology.
Recommender Systems Aalap Kohojkar Yang Liu Zhan Shi March 31, 2008.
Information Retrieval in Practice
Search Engines and Information Retrieval
T.Sharon - A.Frank 1 Internet Resources Discovery (IRD) Classic Information Retrieval (IR)
Basic IR: Queries Query is statement of user’s information need. Index is designed to map queries to likely to be relevant documents. Query type, content,
PROBLEM BEING ATTEMPTED Privacy -Enhancing Personalized Web Search Based on:  User's Existing Private Data Browsing History s Recent Documents 
Time-dependent Similarity Measure of Queries Using Historical Click- through Data Qiankun Zhao*, Steven C. H. Hoi*, Tie-Yan Liu, et al. Presented by: Tie-Yan.
Mobile Web Search Personalization Kapil Goenka. Outline Introduction & Background Methodology Evaluation Future Work Conclusion.
Recommender systems Ram Akella February 23, 2011 Lecture 6b, i290 & 280I University of California at Berkeley Silicon Valley Center/SC.
Recommender Systems; Social Information Filtering.
1 MARG-DARSHAK: A Scrapbook on Web Search engines allow the users to enter keywords relating to a topic and retrieve information about internet sites (URLs)
Recommender systems Ram Akella November 26 th 2008.
Internet Resources Discovery (IRD) Advanced Topics.
Information Retrieval
12 -1 Lecture 12 User Modeling Topics –Basics –Example User Model –Construction of User Models –Updating of User Models –Applications.
Personalized Ontologies for Web Search and Caching Susan Gauch Information and Telecommunications Technology Center Electrical Engineering and Computer.
Overview of Web Data Mining and Applications Part I
Overview of Search Engines
Semantic Web Technologies Lecture # 2 Faculty of Computer Science, IBA.
«Tag-based Social Interest Discovery» Proceedings of the 17th International World Wide Web Conference (WWW2008) Xin Li, Lei Guo, Yihong Zhao Yahoo! Inc.,
Challenges in Information Retrieval and Language Modeling Michael Shepherd Dalhousie University Halifax, NS Canada.
The Exchange of Retrieval Knowledge about Services between Agents Mirjam Minor Mike Wernicke.
Search Engines and Information Retrieval Chapter 1.
Adaptive News Access Daniel Billsus Presented by Chirayu Wongchokprasitti.
Chapter 7 Web Content Mining Xxxxxx. Introduction Web-content mining techniques are used to discover useful information from content on the web – textual.
Dr. Susan Gauch When is a rock not a rock? Conceptual Approaches to Personalized Search and Recommendations Nov. 8, 2011 TResNet.
Xiaoying Gao Computer Science Victoria University of Wellington Intelligent Agents COMP 423.
Recommendation system MOPSI project KAROL WAGA
Chapter 2 Architecture of a Search Engine. Search Engine Architecture n A software architecture consists of software components, the interfaces provided.
UOS 1 Ontology Based Personalized Search Zhang Tao The University of Seoul.
ITIS 1210 Introduction to Web-Based Information Systems Chapter 27 How Internet Searching Works.
1 Information Retrieval Acknowledgements: Dr Mounia Lalmas (QMW) Dr Joemon Jose (Glasgow)
Personalized Search Xiao Liu
University of Malta CSA3080: Lecture 4 © Chris Staff 1 of 14 CSA3080: Adaptive Hypertext Systems I Dr. Christopher Staff Department.
Introduction to Digital Libraries hussein suleman uct cs honours 2003.
Collaborative Information Retrieval - Collaborative Filtering systems - Recommender systems - Information Filtering Why do we need CIR? - IR system augmentation.
Search Engine Architecture
Personalized Interaction With Semantic Information Portals Eric Schwarzkopf DFKI
Next Generation Search Engines Ehsun Daroodi 1 Feb, 2003.
How Do We Find Information?. Key Questions  What are we looking for?  How do we find it?  Why is it difficult? “A prudent question is one-half of wisdom”
Harvesting Social Knowledge from Folksonomies Harris Wu, Mohammad Zubair, Kurt Maly, Harvesting social knowledge from folksonomies, Proceedings of the.
Introduction to Information Retrieval Example of information need in the context of the world wide web: “Find all documents containing information on computer.
Information Retrieval
Web Information Retrieval Prof. Alessandro Agostini 1 Context in Web Search Steve Lawrence Speaker: Antonella Delmestri IEEE Data Engineering Bulletin.
Web Search Personalization with Ontological User Profile Advisor: Dr. Jai-Ling Koh Speaker: Shun-hong Sie.
Augmenting (personal) IR Readings Review Evaluation Papers returned & discussed Papers and Projects checkin time.
Personalization Services in CADAL Zhang yin Zhuang Yuting Wu Jiangqin College of Computer Science, Zhejiang University November 19,2006.
Identifying “Best Bet” Web Search Results by Mining Past User Behavior Author: Eugene Agichtein, Zijian Zheng (Microsoft Research) Source: KDD2006 Reporter:
Artificial Intelligence Techniques Internet Applications 4.
GUILLOU Frederic. Outline Introduction Motivations The basic recommendation system First phase : semantic similarities Second phase : communities Application.
University of Malta CSA3080: Lecture 10 © Chris Staff 1 of 18 CSA3080: Adaptive Hypertext Systems I Dr. Christopher Staff Department.
Information Retrieval in Practice
Data Mining: Concepts and Techniques
Search Engine Architecture
Personalized Social Image Recommendation
Information Retrieval on the World Wide Web
Information Retrieval
Author: Kazunari Sugiyama, etc. (WWW2004)
Magnet & /facet Zheng Liang
Search Engine Architecture
Information Retrieval and Web Design
Introduction to Search Engines
Presentation transcript:

Personalised Search on the World Wide Web Originally by Micarelli, Gasparetti, Sciarrone & Gauch

Group: Tiki Jan Vosecky: ⅓ Contribution Jonathan Abbey: ⅓ Contribution James Griffin: ⅓ Contribution

Contents Introduction Searching The World Wide Web Personalised Search Contextual Search Rich User Models Collaborative Search Engines Adaptive Result Clustering Conclusions

Searching The WWW A large, highly dynamic environment Three main modes of interaction: Searching by Surfing Searching by Query Recommendation

Searching by Query Most popular way that users begin seeking information String queries, on average 2-3 terms, sent to a search engine Classic Information Retrieval (IR) model Information Filtering is more appropriate to the WWW, but suffers from computational complexity

Personalised Search Aims to provide individualised collections of pages Motivated by two problems: Information Overload Problem: over- abundance of resources Vocabulary Problem: issues with polysemy (multiple meanings for one word) and synonymy (multiple words for one meaning)

User Model Needs to build a user model, either......explicitly - the system learns by asking the user...implicitly - the system learns by observing the user...or a combination of both Applied......as a part of the retrieval process...through re-ranking the IR results...through modifying the user’s query

Personalisation Approaches Method Techniqu e Typical Input Data Current Contextimplicit documents, s, history Search Historyimplicit past queries, selected results Rich User Modelsboth past queries, user feedback Collaborativebothuser ratings, past queries Result Clusteringexplicitselected clusters Hypertextual Data bothqueries, selected pages

Contextual Search Based on the user’s current context Implicit feedback unobtrusively draws usage data by tracking and monitoring the user Similar to ‘Search History’, but instead draws from what the user is currently working on or in

Contextual Search Just-in-Time IR (JITIR) (Rhodes), suggests information such as related documents based on what the user is currently reading or typing Watson monitors the user’s actions and predicts the user’s needs, using search engines such as Yahoo! and news sources such as Reuters

Rich User Models Build and update a User Model/Profile –Semantic network representation Using explicit user feedback

ifWeb User profile is represented as a weighted semantic network –Nodes correspond to terms (concepts) from documents the user identified as positive or negative relevance feedback –Arcs link pairs of terms that co-occurred in a document High-level operation: –Collection of documents presented to the user, based on their profile –Update the user profile based on the user’s selection from the results –“Rent”: weight of concepts not used decreases with time

Wifs Filtering of search results returned by a search engine User model consists of –Slots, which contain topics –Each slot is associated with other co- keywords Database of topics is built manually by experts Document is represented as –Planets = topics, found in the database –Satellites = co-keywords associated with topics

Wifs Can calculate a weighted occurrence of a term in a document User ModelDocument Model Calculate document relevance by comparing UM with DM Dynamically updating the user model through relevance feedback on viewed documents

InfoWeb Retrieval of digital library documents User modelling through a stereotype knowledge base –Most typical document for a category of users –Represented as a document vector –Defined by a domain expert Document clustering, with stereotypes being initial cluster centres User profile updated by explicit relevance feedback Query expansion based on the user profile

Collaborative Search Engines Improves search results by learning from the experience of a group of users Calculate similarity measures among user needs –Queries ↔ selected documents from results –Relatedness among two queries might not depend on actual query terms, but on the documents returned E.g. ‘handheld devices’ and ‘mobile computers’

Collaborative Search Engines Eurekster.com – search engine with collaborative (social) filtering –Stores all results selected by users for each query –Social ranking: preferences are shared among a community of users with same interests –Individual ranking: individual user preferences also taken into account when ranking results

Collaborative Search Engines I-Spy – collaborative search engine –Analyses interests of communities of users Community = users of a specialised web site –Statistical model of query-page relevance Probability a page p is selected by a user as a result of query q No need for document content analysis

Adaptive Result Clustering Queries often return thousands of results Impossible to sift through Groups results into clusters By topic Exhaustive partition / Hierarchical tree Clusters ranked against query User navigates clusters

Requirements Usually after results retrieved Must be fast Usually take document snippets Set number & organisation of clusters Maximise ease of navigation Cluster descriptions Concise & accurate

Web Clustering CLUSTY Organise into folder structure by topic KARTOO Folder structure Graphic interactive map Mouse-over descriptions Resource icons sized by relevance

Web Clustering SnakeT Users can select most relevant cluster subset Query refined using contained keywords Scatter/Gather Users select cluster(s) for further analysis Re-organises selection to small number of clusters Resources shown when clusters small

Conclusions Possible solution to overloading problem Increased recent interest Personalisation simplifies searching Also recommendation, data gathering Novel and lively field New approaches likely in next few years

Current Work Building ontology-based UMs implicitly Mapping metadata and ontology concepts to UM stored concepts Logical mapping rules & description logics Plan recognition Recognition and predication of goals Language semantic analysis Relation to user queries

Any Questions?