Presentation is loading. Please wait.

Presentation is loading. Please wait.

Light-weight Domain-based Form Assistant: Querying Web Databases On The Fly Authors:Z. Zhang, B. He, K. C.-C. Chang (Univ. of Illinois at Urbana-Champaign)

Similar presentations


Presentation on theme: "Light-weight Domain-based Form Assistant: Querying Web Databases On The Fly Authors:Z. Zhang, B. He, K. C.-C. Chang (Univ. of Illinois at Urbana-Champaign)"— Presentation transcript:

1 Light-weight Domain-based Form Assistant: Querying Web Databases On The Fly Authors:Z. Zhang, B. He, K. C.-C. Chang (Univ. of Illinois at Urbana-Champaign) Published in:Proceedings of the 31 st VLDB Conference, Trondheim, Norway 2005 Presented by: Bruce Vincent CSE-718 Seminar April 25, 2008

2 Outline Overview Problem Description, Motivating Example System Architecture Design Approaches Query Modeling and Translation Dynamic Predicate Mapping Implementation - Form Assistant Toolkit Experiments Related Work

3 Problem Description “Deep Web” Estimated to contain 450,000 online databases (2004) Sometimes referred to as “Invisible Web” or “Hidden Web” Much of this is accessible only by query forms instead of static URL links Common domains such as: books, cars, airfares

4 Problem Description Often it can be useful to query multiple alternative sources in the same domain Automation of this entails several components One key component is dynamic query translation Software toolkit “Form Assistant” designed to provide potential translations of user queries for alternative sources e.g., User-entered Amazon form query automatically translated to potential Barnes & Noble form query

5 Problem Description Goals of query translator: Source-generality Built-in translation must generally cope with new or “unseen” sources Domain-portability Translator must be easily customizable with domain-specific knowledge, and thus deployable for new domains

6 Motivating Example Source query Q s on source form S: (e.g. Amazon) Target query form T: (e.g. Barnes & Noble)

7 Motivating Example Source query Q s on source form S Target query form T Tom Clancy U Query Translation Filter: : σ title contain “red storm” and price 12 Union Query Q t *:

8 System Architecture Form Extractor Source query Q s Target query form QI Attribute Matcher: Syntax-based schema matching Predicate Mapper: Type-based search-driven mapping Query Rewriter: Constraint-based query rewriting Target query Q t * Domain-specific Thesaurus Domain-specific type handlers Form Assistant (FA)

9 Design Approaches Query Modeling Vocabulary and Syntax Query Translation Dynamic Predicate Modeling

10 Query Modeling Vocabulary Predicate templates: { P 1, P 2, P 3, P 4, P 5 } Example: P1P1 P3P3 P4P4 P2P2 P5P5

11 Query Modeling Example Vocabulary (predicate templates) P 1 = [author; contain; $au] P 2 = [title; contain; $ti] P 3 = [subject; contain; $su] P 4 = [isbn; contain; $isbn] P 5 = [price; between; $s, $e] Example Syntax (valid conjunctive forms) F 1 = P 1 P 5 F 2 = P 2 P 5 F 3 = P 3 P 5 F 4 = P 4 P 5 F 5 = P 1 F 6 = P 2 F 7 = P 3 F 8 = P 4

12 Query Modeling Example Vocabulary Instantiations p 1 = [author; contain; Tom Clancy] p 2 = [title; contain; red storm] p 5 1 = [price; between; 0-25] p 5 2 = [price; between; 25-45] Corresponding Form Queries: f 1 = p 1 p 5 1 f 2 = p 1 p 5 2 Resultant Union Query: Q t = f 1 f 2

13 Query Modeling Syntax Valid combination of predicate templates {F 1, F 2, F 3, F 4, F 5, F 6, F 7, F 8 } Example (‘v’ indicates ‘valid’): F1F1 F2F2 F3F3 F4F4 F5F5 F6F6 F7F7 F8F8 P 1 (author) νν P 2 (title) νν P 3 (subject) νν P 4 (isbn) νν P 5 (price) vvvv Tom Clancy F1:F1: F2:F2:

14 Query Translation Based on semantic closeness of query predicates: Finds minimal subsuming C min Benefits of this approach: No false positives Minimizes false negatives Has clear semantics, independent of DB content Modular translation

15 Query Translation Example: t1:t1: 0 t2:t2: 2545 s: 350 t1 v t2:t1 v t2: 045 t3:t3: 6545 t1 v t2 v t3:t1 v t2 v t3: 065 ?  C min 25

16 Query Translation Definition: Given source query Q s and target query form T, a query Q t * is a “minimal subsuming translation” w.r.t. T if: 1. Q t * is a validquery w.r.t T 2. Q t * subsumes Q s i.e., for any database instance D i, Q s (D i ) ≤ Q t * (D i ) 3. Q t * is minimal i.e., there is no query Q t such that Q t satisfies (1.) and (2.) above and Q t * subsumes Q t

17 Query Translation Example: Consider source query Q s in first example and three target queries Q t1,Q t2,Q t3 Q t1 and Q t3 subsume Q s while Q t2 does not  Misses price range 0-25  Thus can’t be the best translation C min Prune Q t3 because it subsumes Q t1 That leaves Q t1 as C min Q t1 = (f 1 : p 1 p 5 1 ) (f 2 : p 1 p 5 2 ) Q t2 = f 2 Q t3 = f 3 : p 1 p 1 = [author; contain; Tom Clancy] p 5 1 = [price; between; 0-25] p 5 2 = [price; between; 25-45]

18 Dynamic Predicate Mapping Tasks: Choose operator Fill in values Objective: Minimal subsuming between source and target

19 Dynamic Predicate Mapping Example: Predicate Mapping U Input: output:

20 System Architecture (reminder) Form Extractor Source query Q s Target query form QI Attribute Matcher: Syntax-based schema matching Predicate Mapper: Type-based search-driven mapping Query Rewriter: Constraint-based query rewriting Target query Q t * Domain-specific Thesaurus Domain-specific type handlers Form Assistant (FA)

21 Implementation – Form Assistant Toolkit Form Extractor Parses HTML into query predicate templates [attr; op; val] Details discussed in a different paper [3.] by same research group Attribute Matcher (1:1) Identifies semantically corresponding attributes between forms Customized with domain thesaurus (indexes synonyms for commonly used concepts) Stems (e.g., “children” -> “child) and removes stop words (e.g., “the”) Matched by value type and synonym attributes Predicate Mapper (discussed in previous slides) Query Rewriter Well-studied problem to find minimal subsuming query of given predicate- mapped query (uses approach of [5.] by Papakonstantinou, et al)

22 Experiments Datasets 447 Deep Web sources (query forms) in 8 domains 3 “Basic” domains – each with custom thesaurus in FA  Books, Airfares, Automobiles 5 “New” domains (for tests, these don’t have thesaurus)  Car Rentals, Jobs, Hotels, Movies, Music/Records Test Approach Run the FA to translate 120 form queries Each translation test corresponds to random pairing of sources within a domain Count correct mappings in translation suggested by FA Indicates amount of user effort the Form Assistant has saved

23 Experiments Results: Accuracy Distributions X: % correct predicate translations; Y: % tested query forms Forms with all 1:1 mappings had 87% perfect accuracy for Basic dataset, 85% perfect for New dataset (good domain flexibility) Forms having complex mapping: 76%, 70% “near perfect” (Y>80%) FA did not attempt complex (n:m) mappings, such as a full name in source mapping to separate first and last names in target For Basic dataset:For New dataset:

24 Experiments Accuracy ratio: correct results per 1:1 query Raw: includes some forms whose input form extraction step had errors Perfect: manually forces all correct form extractions Avg. accuracy improves for perfectly correct extraction step: for Basic dataset, 90.4% improves to 96.1% For New dataset, 81.1% improves to 86.7% Basic: 3 domainsNew: 5 domains

25 Experiments Example Error in Form Extraction delta.com form has link to alternative reservation page “One-way & multi-city reservations”  Wrongly interpreted by Form Extractor as input field label (attribute)

26 Experiments Error Distribution % of errors caused by each component Fewest errors are due to Attribute Matching Most errors due to Predicate Mapping Cited reason for PM errors is insufficient domain knowledge  Example failure: source subject value “computer science” didn’t properly map to target subject value “programming languages”  Improvement could entail better domain-specific ontology and type handlers Attribute Matching 18% 40% 42% Form Extraction Predicate Mapping

27 Related Work From the same research group: Complex Matchings (n:m) Defines “Type Recognizer” used in Form Assistant’s Attribute Matcher, and discusses complex n:m matchings not attempted by Form Assistant:  [1.] Discovering Complex Matchings across Web Query Interfaces: A Correlation Mining Approach. B. He, K. C.-C. Chang, and J. Han. In Proceedings of the 2004 ACM SIGKDD Conference (KDD 2004) (Full Paper), Seattle, Washington, August 2004 MetaQuerier System Fuller system for both exploring (to find) and integrating (to query) Deep Web databases:  [2.] Toward Large Scale Integration: Building a MetaQuerier over Databases on the Web. K. C.-C. Chang, B. He, and Z. Zhang. In Proceedings of the Second Conference on Innovative Data Systems Research (CIDR 2005), Asilomar, California, January 2005

28 Related Work From the same research group: Form Extraction As used by implementation of Form Assistant:  [3.] Understanding Web Query Interfaces: Best-Effort Parsing with Hidden Syntax. Z. Zhang, B. He, and K. C.-C. Chang. In Proceedings of the 2004 ACM SIGMOD Conference (SIGMOD 2004), Paris, France, June 2004 2007 thorough analysis of the Deep Web Interesting survey of web databases and query interfaces:  [4.] Accessing the Deep Web: A Survey. B. He, M. Patel, Z. Zhang, and K. C.-C. Chang. Communications of the ACM (CACM), 50(5):94-101, May 2007 Public Datasets Cached real world query form web pages (used in experiments):  http://metaquerier.cs.uiuc.edu/repository/datasets/tel-8 http://metaquerier.cs.uiuc.edu/repository/datasets/tel-8 Additional Deep Web integration resources:  http://metaquerier.cs.uiuc.edu/repository

29 Related Work Query Rewriting As used by implementation of Form Assistant: [5.] Y. Papakonstaninou, A. Gupta, H. Garcia-Molina, and J. Ullman. A query translation scheme for rapid implementation of wrappers In proceedings of the Fourth International Conference on Deductive and Object-Oriented Databases, Singapore, December 1995.

30 Thank you !


Download ppt "Light-weight Domain-based Form Assistant: Querying Web Databases On The Fly Authors:Z. Zhang, B. He, K. C.-C. Chang (Univ. of Illinois at Urbana-Champaign)"

Similar presentations


Ads by Google