Evaluating the Robustness of Learning from Implicit Feedback Filip Radlinski Thorsten Joachims Presentation by Dinesh Bhirud

Slides:



Advertisements
Similar presentations
Answering Approximate Queries over Autonomous Web Databases Xiangfu Meng, Z. M. Ma, and Li Yan College of Information Science and Engineering, Northeastern.
Advertisements

Accurately Interpreting Clickthrough Data as Implicit Feedback Joachims, Granka, Pan, Hembrooke, Gay Paper Presentation: Vinay Goel 10/27/05.
Pseudo-Relevance Feedback For Multimedia Retrieval By Rong Yan, Alexander G. and Rong Jin Mwangi S. Kariuki
Text Categorization.
Improvements and extras Paul Thomas CSIRO. Overview of the lectures 1.Introduction to information retrieval (IR) 2.Ranked retrieval 3.Probabilistic retrieval.
CSE3201/4500 Information Retrieval Systems
Application of Ensemble Models in Web Ranking
Language Models Naama Kraus (Modified by Amit Gross) Slides are based on Introduction to Information Retrieval Book by Manning, Raghavan and Schütze.
1 Evaluation Rong Jin. 2 Evaluation  Evaluation is key to building effective and efficient search engines usually carried out in controlled experiments.
Modelling Relevance and User Behaviour in Sponsored Search using Click-Data Adarsh Prasad, IIT Delhi Advisors: Dinesh Govindaraj SVN Vishwanathan* Group:
Optimizing search engines using clickthrough data
Query Chains: Learning to Rank from Implicit Feedback Paper Authors: Filip Radlinski Thorsten Joachims Presented By: Steven Carr.
Search Engines Information Retrieval in Practice All slides ©Addison Wesley, 2008.
Eye Tracking Analysis of User Behavior in WWW Search Laura Granka Thorsten Joachims Geri Gay.
Query Dependent Pseudo-Relevance Feedback based on Wikipedia SIGIR ‘09 Advisor: Dr. Koh Jia-Ling Speaker: Lin, Yi-Jhen Date: 2010/01/24 1.
Information Retrieval Models: Probabilistic Models
1 Learning User Interaction Models for Predicting Web Search Result Preferences Eugene Agichtein Eric Brill Susan Dumais Robert Ragno Microsoft Research.
Content Based Image Clustering and Image Retrieval Using Multiple Instance Learning Using Multiple Instance Learning Xin Chen Advisor: Chengcui Zhang Department.
Evaluating Search Engine
Learning to Rank: New Techniques and Applications Martin Szummer Microsoft Research Cambridge, UK.
Query Operations: Automatic Local Analysis. Introduction Difficulty of formulating user queries –Insufficient knowledge of the collection –Insufficient.
1 CS 430 / INFO 430 Information Retrieval Lecture 8 Query Refinement: Relevance Feedback Information Filtering.
DYNAMIC ELEMENT RETRIEVAL IN A STRUCTURED ENVIRONMENT MAYURI UMRANIKAR.
1 Statistical correlation analysis in image retrieval Reporter : Erica Li 2004/9/30.
Probabilistic IR Models Based on probability theory Basic idea : Given a document d and a query q, Estimate the likelihood of d being relevant for the.
Modern Information Retrieval Chapter 5 Query Operations.
Topic-Sensitive PageRank Taher H. Haveliwala. PageRank Importance is propagated A global ranking vector is pre-computed.
An investigation of query expansion terms Gheorghe Muresan Rutgers University, School of Communication, Information and Library Science 4 Huntington St.,
Overview of Search Engines
Modeling (Chap. 2) Modern Information Retrieval Spring 2000.
1 Context-Aware Search Personalization with Concept Preference CIKM’11 Advisor : Jia Ling, Koh Speaker : SHENG HONG, CHUNG.
A Comparative Study of Search Result Diversification Methods Wei Zheng and Hui Fang University of Delaware, Newark DE 19716, USA
Improving Web Search Ranking by Incorporating User Behavior Information Eugene Agichtein Eric Brill Susan Dumais Microsoft Research.
UOS 1 Ontology Based Personalized Search Zhang Tao The University of Seoul.
CIKM’09 Date:2010/8/24 Advisor: Dr. Koh, Jia-Ling Speaker: Lin, Yi-Jhen 1.
Karthik Raman, Pannaga Shivaswamy & Thorsten Joachims Cornell University 1.
Giorgos Giannopoulos (IMIS/”Athena” R.C and NTU Athens, Greece) Theodore Dalamagas (IMIS/”Athena” R.C., Greece) Timos Sellis (IMIS/”Athena” R.C and NTU.
Implicit User Feedback Hongning Wang Explicit relevance feedback 2 Updated query Feedback Judgments: d 1 + d 2 - d 3 + … d k -... Query User judgment.
Context-Sensitive Information Retrieval Using Implicit Feedback Xuehua Shen : department of Computer Science University of Illinois at Urbana-Champaign.
Machine Learning in Ad-hoc IR. Machine Learning for ad hoc IR We’ve looked at methods for ranking documents in IR using factors like –Cosine similarity,
Search Engines that Learn from Implicit Feedback Jiawen, Liu Speech Lab, CSIE National Taiwan Normal University Reference: Search Engines that Learn from.
1 Computing Relevance, Similarity: The Vector Space Model.
Distributed Information Retrieval Server Ranking for Distributed Text Retrieval Systems on the Internet B. Yuwono and D. Lee Siemens TREC-4 Report: Further.
Lecture 1: Overview of IR Maya Ramanath. Who hasn’t used Google? Why did Google return these results first ? Can we improve on it? Is this a good result.
IR Theory: Relevance Feedback. Relevance Feedback: Example  Initial Results Search Engine2.
 Examine two basic sources for implicit relevance feedback on the segment level for search personalization. Eye tracking Display time.
This paper was presented at KDD ‘06 Discovering Interesting Patterns Through User’s Interactive Feedback Dong Xin Xuehua Shen Qiaozhu Mei Jiawei Han Presented.
LANGUAGE MODELS FOR RELEVANCE FEEDBACK Lee Won Hee.
Personalizing Web Search using Long Term Browsing History Nicolaas Matthijs, Cambridge Filip Radlinski, Microsoft In Proceedings of WSDM
Carnegie Mellon Novelty and Redundancy Detection in Adaptive Filtering Yi Zhang, Jamie Callan, Thomas Minka Carnegie Mellon University {yiz, callan,
Implicit User Feedback Hongning Wang Explicit relevance feedback 2 Updated query Feedback Judgments: d 1 + d 2 - d 3 + … d k -... Query User judgment.
Learning to Estimate Query Difficulty Including Applications to Missing Content Detection and Distributed Information Retrieval Elad Yom-Tov, Shai Fine,
A Framework to Predict the Quality of Answers with Non-Textual Features Jiwoon Jeon, W. Bruce Croft(University of Massachusetts-Amherst) Joon Ho Lee (Soongsil.
The Effect of Database Size Distribution on Resource Selection Algorithms Luo Si and Jamie Callan School of Computer Science Carnegie Mellon University.
1 CS 430 / INFO 430 Information Retrieval Lecture 12 Query Refinement and Relevance Feedback.
To Personalize or Not to Personalize: Modeling Queries with Variation in User Intent Presented by Jaime Teevan, Susan T. Dumais, Daniel J. Liebling Microsoft.
Usefulness of Quality Click- through Data for Training Craig Macdonald, ladh Ounis Department of Computing Science University of Glasgow, Scotland, UK.
Autumn Web Information retrieval (Web IR) Handout #14: Ranking Based on Click Through data Ali Mohammad Zareh Bidoki ECE Department, Yazd University.
Information Retrieval in Practice
Accurately Interpreting Clickthrough Data as Implicit Feedback
Large-Scale Content-Based Audio Retrieval from Text Queries
Lecture 12: Relevance Feedback & Query Expansion - II
Evaluation of IR Systems
Martin Rajman, Martin Vesely
Toshiyuki Shimizu (Kyoto University)
Authors: Wai Lam and Kon Fan Low Announcer: Kyu-Baek Hwang
CS246: Leveraging User Feedback
How does Clickthrough Data Reflect Retrieval Quality?
INF 141: Information Retrieval
SVMs for Document Ranking
Presentation transcript:

Evaluating the Robustness of Learning from Implicit Feedback Filip Radlinski Thorsten Joachims Presentation by Dinesh Bhirud

Introduction The paper evaluates the robustness of learning to rank documents based on Implicit feedback. What is implicit feedback? – Relevance feedback obtained from search engine log files – Easier to collect large amount of such training data as against explicitly collecting relevance feedback.

Osmot Osmot – Search engine developed at Cornell University based on Implicit Feedback Name Osmot comes from the word osmosis – learning from the users by osmosis Query Chains – Sequence of reformulated queries. – Osmot learns ranked retrieval function by observing query chains and monitoring user clicks.

High Level Block Diagram Data generation User behavior simulation (based on original ranking fucntion) Preference generation SVM Learning User behavior simulatoin (based on learned ranking function)

Data Generation Set of W words are chosen, word frequencies obeying a Ziphs law T topics are picked by picking N words/topic uniformly from W. Each document d is generated as – Pick k d binomially from [0,T] – Repeat k d times Pick topic t Pick L/k d words from topic t.

Relevance 3 kinds of relevance – Relevance with respect to topic Can be measured/known because document collection and topics are synthetic Used for evaluating the ranking function. – Relevance with respect to query Actual relevance score of a document with respect to a query Used to rank documents – Observed relevance Relevance of a document as judged by the user seeing only the abstract. Used to simulate user behavior.

User behavior parameters Noise – Accuracy of users relevance estimate – Affects observed relevance. (obsRel) – obsRel is drawn from an incomplete Beta distribution where α gives noise level and β is selected so that mode is at rel(d,q) Threshold – User selectivity over results (rT) Patience – Number of results user looks at before giving up (rP) Reformulation – How likely is the user to reformulate query(P reform )

User Behavior Model While question T is unanswered 1.1 Generate query q (Let d1,d2..,dn be results for q) 1.2 Start with document 1 ie i = while patience (Rp) > if obsRel(di,q) > rT if obsRel(di+1, q) > obsRel(di,q) + c then continue looking further in the list else di is a good document, click on it. If rel(di,T) is 1, user is DONE Decrease patience Rp else Decrease patience Rp Rp = Rp - (rT – obsRel(di,q)) Set i = i With probability (1 – Preform), user gives up.

User Preference Model Based on the clickthrough log files, users preferences for documents given query q can be found. Clickthrough logs generated by simulating users. From preference, features values are calculated.

Feedback Strategies Single Query Strategy Click > q Skip Above – For query q, if document d i is clicked, d i is preferred over all d j, j < i. Click 1 st > q No-Click 2 nd – For query q, if document 1 is clicked, it is preferred over the 2 nd document in the list.

Feedback Strategies 2-Query Strategy 1 This strategy uses 2 queries in a query chain, but document rankings only for the later query. Given queries q' and q in a query chain Click > q' Skip Above – For query q', if document di is clicked in query q, di is preferred over all dj, j < i Click 1 st > q' No-Click 2 nd – For query q', if document 1 is clicked, it is preferred over the 2 nd document in the list for q

Feedback Strategies 2-Query Strategy 2 This strategy uses 2 queries in a query chain, and document rankings for both used. Given queries q' and q in a query chain Click > q' Skip Earlier Query – For query q', if document di is clicked in query q, di is preferred over seen documents in query previous query. Click > q' Top two earlier Query – If no document clicked for query q', then di preferred over top two in previous query.

Example Q1Q2 D1D4 D2D5 D3D6 Preferences D2 > q1 D1 D4 > q2 D5 D4 > q1 D5 D4 > q1 D1 D4 > q1 D3

Features Document d i would be mapped to feature vector with respect to query q. 2 types of features defined – Rank Features – Term/Document Features

Rank Features Rank features allow representation of ranking given by the existing static retrieval function. Used a simple TFIDF weighted cosine similarity metric (rel 0 ) 28 rank features used for ranks 1,2,..,10,15,20,…100. Set to 1 if clicked document is at or above specified rank.

Term Features Allows representation of fine grained relationship between query terms and documents. If for query q, document d is clicked, then for each word, Forms a sparse feature vector, as only very few words are included in query.

Learning Retrieval Function rel(d i, q) defined as where is the weight vector. Intuitively, weight vector assigns weight to each feature identified. Task of learning a ranking function is reduced to the task of learning an optimal weight vector.

How does affect ranking? Points are ordered by their projections onto For the ordering will be 1,2,3,4. For the ordering will be 2,3,1,4. Weight vector needs to be learnt that will minimize number of discordant rankings.

Learning Problem Learning problem can be formalized as follows Find weight vector such that maximum of following inequalities fulfilled. such that then Without using slack variables, this is NP-hard problem.

SVM Learning Equivalent optimization problem would be Minimize Subject to rearranging which we get constraint and

Re-ranking using the learnt model SVM-Light package is used. Model provides values for all support vectors. User behavior is again simulated, this time using the learnt ranking function. How does reranking work? – First, a ranked list of documents is obtained using the original ranking function. – This list is re-ordered, using the weights of each feature obtained from the learnt model.

Experiments Experiments done to study the behavior of the search engine by varying parameters like – Noise in users relevance judgement – Ambiguity of words in topics and queries – Threshold value which user considers good document – Users trust in ranking – Users probability of reformulation of query.

Results - Noise

Noise – My experiment Did implementation for extracting preferences and encoding them in features.

Topic and Word Ambiguity

Probability of user reformulating query

Thank You