Presentation is loading. Please wait.

Presentation is loading. Please wait.

DASFAA 2003BYU Data Extraction Group Discovering Direct and Indirect Matches for Schema Elements Li Xu and David W. Embley Brigham Young University Funded.

Similar presentations


Presentation on theme: "DASFAA 2003BYU Data Extraction Group Discovering Direct and Indirect Matches for Schema Elements Li Xu and David W. Embley Brigham Young University Funded."— Presentation transcript:

1 DASFAA 2003BYU Data Extraction Group Discovering Direct and Indirect Matches for Schema Elements Li Xu and David W. Embley Brigham Young University Funded by NSF

2 DASFAA 2003BYU Data Extraction Group Information Exchange SourceTarget Information Extraction Schema Matching Leverage this … … to do this

3 DASFAA 2003BYU Data Extraction Group Outline Information Extraction Direct Schema Matching Indirect Schema Matching Schema Matching for HTML Tables Conclusions

4 DASFAA 2003BYU Data Extraction Group Outline Information Extraction Direct Schema Matching Indirect Schema Matching Schema Matching for HTML Tables Conclusions

5 DASFAA 2003BYU Data Extraction Group Extracting Pertinent Information from Documents

6 DASFAA 2003BYU Data Extraction Group A Conceptual-Modeling Solution YearPrice Make Mileage Model Feature PhoneNr Extension Car has is for has 1..* 0..1 1..* 0..1 0..* 1..*

7 DASFAA 2003BYU Data Extraction Group Car-Ads Ontology Car [->object]; Car [0..1] has Year [1..*]; Car [0..1] has Make [1..*]; Car [0...1] has Model [1..*]; Car [0..1] has Mileage [1..*]; Car [0..*] has Feature [1..*]; Car [0..1] has Price [1..*]; PhoneNr [1..*] is for Car [0..*]; PhoneNr [0..1] has Extension [1..*]; Year matches [4] constant {extract “\d{2}”; context "([^\$\d]|^)[4-9]\d[^\d]"; substitute "^" -> "19"; }, … End;

8 DASFAA 2003BYU Data Extraction Group Recognition and Extraction Car Year Make Model Mileage Price PhoneNr 0001 1989 Subaru SW $1900 (336)835-8597 0002 1998 Elantra (336)526-5444 0003 1994 HONDA ACCORD EX 100K (336)526-1081 Car Feature 0001 Auto 0001 AC 0002 Black 0002 4 door 0002 tinted windows 0002 Auto 0002 pb 0002 ps 0002 cruise 0002 am/fm 0002 cassette stereo 0002 a/c 0003 Auto 0003 jade green 0003 gold

9 DASFAA 2003BYU Data Extraction Group Outline Information Extraction Direct Schema Matching Indirect Schema Matching Schema Matching for HTML Tables Conclusions

10 DASFAA 2003BYU Data Extraction Group Attribute Matching for Populated Schemas Central Idea: Exploit All Data & Metadata Matching Possibilities (Facets) –Attribute Names –Data-Value Characteristics –Expected Data Values –Data-Dictionary Information –Structural Properties

11 DASFAA 2003BYU Data Extraction Group Approach Target Schema T Source Schema S Framework –Individual Facet Matching –Combining Facets –Best-First Match Iteration

12 DASFAA 2003BYU Data Extraction Group Example Source Schema S Car Year has 0:1 Make has 0:1 Model has 0:1 Cost Style has 0:1 0:* Year has 0:1 Feature has 0:* Cost has 0:1 Car Mileage has Phone has 0:1 Model has 0:1 Target Schema T Make has 0:1 Miles has 0:1 Year Model Make Year Make Model Car MileageMiles

13 DASFAA 2003BYU Data Extraction Group Individual Facet Matching Attribute Names Data-Value Characteristics Expected Data Values

14 DASFAA 2003BYU Data Extraction Group Attribute Names Target and Source Attributes –T : A –S : B WordNet C4.5 Decision Tree: feature selection, trained on schemas in DB books –f0: same word –f1: synonym –f2: sum of distances to a common hypernym root –f3: number of different common hypernym roots –f4: sum of the number of senses of A and B

15 DASFAA 2003BYU Data Extraction Group WordNet Rule The number of different common hypernym roots of A and B The sum of distances of A and B to a common hypernym The sum of the number of senses of A and B

16 DASFAA 2003BYU Data Extraction Group Confidence Measures

17 DASFAA 2003BYU Data Extraction Group Data-Value Characteristics C4.5 Decision Tree Features –Numeric data (Mean, variation, standard deviation, …) –Alphanumeric data (String length, numeric ratio, space ratio)

18 DASFAA 2003BYU Data Extraction Group Confidence Measures

19 DASFAA 2003BYU Data Extraction Group Expected Data Values Target Schema T and Source Schema S –Regular expression recognizer for attribute A in T –Data instances for attribute B in S Hit Ratio = N'/N for (A, B) match –N' : number of B data instances recognized by the regular expressions of A –N: number of B data instances

20 DASFAA 2003BYU Data Extraction Group Confidence Measures

21 DASFAA 2003BYU Data Extraction Group Combined Measures Threshold: 0.5 1 0 0 0 0 0 0 0 000000 1 0 0 0 0 0 0000 10 0 0000 0 0 0 0 0 1 0 0 0 00 100 0 0 0 0 0

22 DASFAA 2003BYU Data Extraction Group Final Confidence Measures 0 0 0

23 DASFAA 2003BYU Data Extraction Group Experimental Results This schema, plus 6 other schemas –32 matched attributes –376 unmatched attributes Matched: 100% Unmatched: 99.5% –“Feature” ---”Color” –“Feature” ---”Body Type” F1 93.8% F2 84% F3 92% F1 98.9% F2 97.9% F3 98.4% F1: WordNet F2: Value Characteristics F3: Expected Values

24 DASFAA 2003BYU Data Extraction Group Outline Information Extraction Direct Schema Matching Indirect Schema Matching Schema Matching for HTML Tables Conclusions

25 DASFAA 2003BYU Data Extraction Group Schema Matching Source Car Year Cost Style Year Feature Cost Phone Target Car Miles Mileage Model Make & Model Color Body Type

26 DASFAA 2003BYU Data Extraction Group Mapping Generation Direct Matches as described earlier: –Attribute Names based on WordNet –Value Characteristics based on value lengths, averages, … –Expected Values based on regular-expression recognizers Indirect Matches: –Direct matches –Structure Evaluation Union Selection Decomposition Composition

27 DASFAA 2003BYU Data Extraction Group Union and Selection Car Source Year Cost Style Year Feature Cost Phone Target Car Miles Mileage Model Make & Model Color Body Type

28 DASFAA 2003BYU Data Extraction Group Decomposition and Composition Car Source Year Cost Style Year Feature Cost Phone Target Car Miles Mileage Model Make & Model Color Body Type

29 DASFAA 2003BYU Data Extraction Group Structure PO POShipToPOBillToPOLines CityStreetCityStreetItem Count LineQtyUoM PurchaseOrder DeliverToInvoiceTo Items ItemItemCount ItemNumber QuantityUnitOfMeasure CityStreet Address TargetSource Example Taken From [MBR, VLDB’01]

30 DASFAA 2003BYU Data Extraction Group Structure (Nonlexical Matches) PO POShipToPOBillToPOLines CityStreetCityStreetItem Count LineQtyUoM PurchaseOrder DeliverToInvoiceTo Items ItemCount ItemNumber QuantityUnitOfMeasure CityStreet Address DeliverTo TargetSource

31 DASFAA 2003BYU Data Extraction Group Structure (Join over FD Relationship Sets, …) PO POBillToPOLines CityStreetCityStreetItem Count LineQtyUoM PurchaseOrder InvoiceTo Items ItemCount ItemNumber QuantityUnitOfMeasure City Street City Street POShipToDeliverTo TargetSource

32 DASFAA 2003BYU Data Extraction Group Structure (Lexical Matches) PO POBillToPOLines CityStreetCityStreetItem Count LineQtyUoM PurchaseOrder InvoiceTo Items ItemCount ItemNumber Quantity City Street City Street City Street City Street Count LineQty QuantityUnitOfMeasure POShipToDeliverTo TargetSource

33 DASFAA 2003BYU Data Extraction Group Experimental Results Applications (Number of Schemes) Precision (%) Recall (%) F (%) CorrectFalse Positive False Negative Course Schedule (5) 98939611929 Faculty Member (5) 100 14000 Real Estate (5) 9296942352010 Data borrowed from Univ. of Washington [DDH, SIGMOD01] Indirect Matches: 94% (precision, recall, F-measure) Rough Comparison with U of W Results (direct matches only) * Course Schedule – Accuracy: ~71% * Faculty Members – Accuracy, ~92% * Real Estate (2 tests) – Accuracy: ~75%

34 DASFAA 2003BYU Data Extraction Group Outline Information Extraction Direct Schema Matching Indirect Schema Matching Schema Matching for HTML Tables Conclusions

35 DASFAA 2003BYU Data Extraction Group Problem: Different Schemas Target Database Schema {Car, Year, Make, Model, Mileage, Price, PhoneNr}, {PhoneNr, Extension}, {Car, Feature} Different Source Table Schemas –{Run #, Yr, Make, Model, Tran, Color, Dr} –{Make, Model, Year, Colour, Price, Auto, Air Cond., AM/FM, CD} –{Vehicle, Distance, Price, Mileage} –{Year, Make, Model, Trim, Invoice/Retail, Engine, Fuel Economy}

36 DASFAA 2003BYU Data Extraction Group Solution Form attribute-value pairs Do extraction Infer mappings from extraction patterns

37 DASFAA 2003BYU Data Extraction Group Solution: Remove Internal Factoring Discover Nesting: Make, (Model, (Year, Colour, Price, Auto, Air Cond, AM/FM, CD)*)* Unnest: μ (Model, Year, Colour, Price, Auto, Air Cond, AM/FM, CD)* μ (Year, Colour, Price, Auto, Air Cond, AM/FM, CD)* Table Legend ACURA

38 DASFAA 2003BYU Data Extraction Group Solution: Replace Boolean Values Legend ACURA β CD Table Yes, CD Yes, β Auto β Air Cond β AM/FM Yes, AM/FM Air Cond. Auto

39 DASFAA 2003BYU Data Extraction Group Solution: Form Attribute-Value Pairs Legend ACURA CD AM/FM Air Cond. Auto,,,,,,,,

40 DASFAA 2003BYU Data Extraction Group Solution: Do Extraction Legend ACURA CD AM/FM Air Cond. Auto

41 DASFAA 2003BYU Data Extraction Group Solution: Infer Mappings Legend ACURA CD AM/FM Air Cond. Auto {Car, Year, Make, Model, Mileage, Price, PhoneNr}, {PhoneNr, Extension}, {Car, Feature} Each row is a car. π Model μ (Year, Colour, Price, Auto, Air Cond, AM/FM, CD)* Table π Make μ (Model, Year, Colour, Price, Auto, Air Cond, AM/FM, CD)* μ (Year, Colour, Price, Auto, Air Cond, AM/FM, CD)* Table π Year Table Note: Mappings produce sets for attributes. Joining to form records is trivial because we have OIDs for table rows (e.g. for each Car).

42 DASFAA 2003BYU Data Extraction Group Solution: Do Extraction Legend ACURA CD AM/FM Air Cond. Auto {Car, Year, Make, Model, Mileage, Price, PhoneNr}, {PhoneNr, Extension}, {Car, Feature} π Model μ (Year, Colour, Price, Auto, Air Cond, AM/FM, CD)* Table

43 DASFAA 2003BYU Data Extraction Group Solution: Do Extraction Legend ACURA CD AM/FM Air Cond. Auto {Car, Year, Make, Model, Mileage, Price, PhoneNr}, {PhoneNr, Extension}, {Car, Feature} π Price Table

44 DASFAA 2003BYU Data Extraction Group Solution: Do Extraction Legend ACURA CD AM/FM Air Cond. Auto {Car, Year, Make, Model, Mileage, Price, PhoneNr}, {PhoneNr, Extension}, {Car, Feature} Yes, ρ Colour←Feature π Colour Table U ρ Auto ← Feature π Auto β Auto Table U ρ Air Cond. ← Feature π Air Cond. β Air Cond. Table U ρ AM/FM ← Feature π AM/FM β AM/FM Table U ρ CD ← Feature π CD β CD Table Yes,

45 DASFAA 2003BYU Data Extraction Group Experiment Tables from 60 sites 10 “ training ” tables 50 test tables 357 mappings (from all 60 sites) –172 direct mappings (same attribute and meaning) –185 indirect mappings (29 attribute synonyms, 5 “ Yes/No ” columns, 68 unions over columns for Feature, 19 factored values, and 89 columns of merged values that needed to be split)

46 DASFAA 2003BYU Data Extraction Group Results 10 “training” tables –100% of the 57 mappings –No false mappings 50 test tables –94.7% of the 300 mappings –No false mappings 16 missed mappings –4 partial (not all unions included) –6 non-U.S. car-ads (unrecognized makes and models) –2 U.S. unrecognized makes and models –3 prices (missing $ or found MSRP instead) –1 mileage (mileages less than 1,000)

47 DASFAA 2003BYU Data Extraction Group Conclusions Direct Attribute Matching –Matched 32 of 32: 100% Recall –2 False Positives: 94% Precision Direct and Indirect Attribute Matching –Matched 494 of 513: 96% Recall –22 False Positives: 96% Precision Table Mappings –Matched 284 of 300: 94.7% Recall –No False Positives: 100% Precision www.deg.byu.edu


Download ppt "DASFAA 2003BYU Data Extraction Group Discovering Direct and Indirect Matches for Schema Elements Li Xu and David W. Embley Brigham Young University Funded."

Similar presentations


Ads by Google