Collectively Representing Semi-Structured Data from the Web

Bhavana Dalvi , William W. Cohen and Jamie CallanLanguage Technologies Institute

Carnegie Mellon University

Paper ID : 02

1

This work is supported by Google and the Intelligence Advanced Research Projects Activity (IARPA) via Air Force Research Laboratory (AFRL) contract number FA8650-10-C-7058.

Motivation Entities on the Web can be present in multiple datasets. E.g.

HTML tables, text documents etc. Traditional systems : Entities as sparse vector of document Ids

in which it occurs. We propose a low-dimensional representation for such entities. Helps to efficiently perform different tasks with a small number

of primitive operations : Semi-supervised Learning (SSL) Set Expansion (SE) Automatic Class Instance Acquisition (ASIA)

2

3

Entities in HTML tables

TC-2 TC-3

Country Sports

India Hockey

UK Cricket

USA Tennis

Country Capital City

India Delhi

USA Washington DC

Canada Ottawa

France Paris USA

India

Hockey

Cricket

Tennis

TC-1

TC-2

TC-3

TC-4

EntityTable-column

Entity-ColumnBi-partite Graph

4

Entities in unstructured text

USA

India

Hockey

Cricket

Tennis

Country

Location

Sports

SuchasEntity

“Such as”Bi-partite Graph

Countries such as India are developing rapidly in terms of

infrastructure.

Outdoor sports include Tennis and Cricket.

5

Resultant Tri-partite Graph

USA

India

Hockey

Cricket

Tennis

Country

Location

Sports

TC-1

TC-2

TC-3

TC-4

SuchasEntity

Table-column

“Such as”Bi-partite Graph

Entity-ColumnBi-partite Graph

6

Encoding the graph“Entity-Column”Bi-partite Graph

Entity X1 X2

USA 0.43 0.66

India 0.41 0.69

Hockey 0.36 0.80

Cricket 0.35 0.82

Tennis 0.34 0.79

Low-dimensional embedding using bipartite Power Iteration Clustering (Lin & Cohen, ICML 2010/ECAI 2010)

USA

India

Hockey

Cricket

Tennis

TC-1

TC-2

TC-3

TC-4

EntityTable-column

Entities with similar X1/X2 values should be ontologically similar - values summarize tabular co-occurrence

7

Encoding the graph

USA

India

Hockey

Cricket

Tennis

Country

Location

Sports

SuchasEntity

“Such as”Bi-partite Graph

Entity Y1 Y2

USA 0.23 0.76

India 0.21 0.79

Hockey 0.66 0.35

Cricket 0.16 0.92

Tennis 0.14 0.89

Low-dimensional embedding using bipartite Power Iteration Clustering (Lin & Cohen, ICML 2010/ECAI 2010)

Entities with similar Y1/Y2 values should be ontologically similar - values summarize “such as pattern” co-occurrence

Low-dimensional PIC3 embedding

n * t entity-tableColumn

Bipartite graph

n * s entity-suchas Bipartite graph

n * m PIC embeddingm << t

n * m PIC embeddingm << s

n * 2m PIC3 embeddingPIC

PIC

Concatenate

Entity X1 X2

USA 0.43 0.66

India 0.41 0.69

Hockey 0.36 0.80

Cricket 0.35 0.82

Tennis 0.34 0.79

Y1 Y2

0.23 0.76

0.21 0.79

0.66 0.35

0.16 0.92

0.14 0.89

9

Using PIC3 Representation

• Semi-Supervised Learning : Given few seed examples for each class, predict class-labels for unlabeled data-points.

• Set Expansion : Given a set of seed entities, find more entities similar to seed entities.

• Automatic Set Instance Acquisition (ASIA) : Given a concept name automatically find instances of that concept.

Quantitative Evaluation: DatasetsDataset Toy_Apple Delicious_Sports

#entities 14,996 438

# table-columns 156 925#entity-table column edges 176,598 9,192#suchas concepts 2,348 1,649#entity-suchas edges 7,683 4,799#general entity classes (NELL KB) 11 3#entities in general classes 419 39#hand-coded column types 31 30#columns in labeled types 156 925

Link to dataset: http://rtw.ml.cmu.edu/wk/WebSets/wsdm_2012_online

11

Task Training Testing

Semi-Supervised Learning

PIC3 + Train SVM classifier

Predict using learnt SVM model

SSL using PIC3Input : Few seed examples for each class label

Output : Class-labels for unlabeled data-points

PIC clusters similar entities together better SVM classifier on unlabeled data (use of background data)

SSL Task - I

12

# dimensions : 2504 10

SSL Task - II

13

# dimensions : 2574 10

14

Task Training Testing

Set Expansion

PIC3 Centroid(entity set) + K-NN (centroid)

Set Expansion using PIC3Input : Few seed entities e.g. Football, Hockey, Tennis

Output : More entities of same type as seeds e.g. Baseball, Badminton, Cricket, Golf ….

K-NN operation is extremely efficient using KD-trees.

Query Times• PIC3 preprocessing : 0.02 sec• # SE queries = 881

• Precision Recall Curve : K-NN+PIC3 consistently beats K-NN-Baseline. Modified Adsorption method is better on 2/5 query classes at the expense of larger query time.

15

Method Total Query Time (s)K-NN + PIC3 12.7 K-NN-Baseline 80.1 MAD 38.2

Modified Adsorption : Graph based label

propagation algorithm

16

Task Training Testing

Automatic Set Instance Acquisition

PIC3 + Inverted index (suchasConcept entities)

seeds = top-k-entities (lookup concept in index)+ Set Expansion (seeds)

Automatic Set Instance Acquisition(ASIA) : using PIC3

Input : Class label e.g. Country

Output : Entities belonging to the given class label e.g. India, China, USA, Canada, Japan …..

Previously described Set Expansion algorithm is used as a subroutine here.

Query Times• PIC3 preprocessing : 0.02 sec• # ASIA queries = 25

• Precision Recall Curve : K-NN+PIC3 consistently beats K-NN-Baseline. Modified Adsorption method is better on 2/4 query classes at the expense of much larger query time.

17

Method Total Query Time (s)K-NN + PIC3 0.5K-NN-Baseline 1.4MAD 150.0

Conclusions & Future Work Presented a novel low-dimensional PIC3 representation for

entities on the Web using Power Iteration Clustering (PIC). Simple primitive operations on PIC3 to perform following tasks :

Semi-Supervised Learning Set Expansion Automatic Set Instance Acquisition

Future work : Use PIC3 representation for Named entity disambiguation and Unsupervised class-instance pair acquisition

18

Thank You !!

19

This work is supported by Google and the Intelligence Advanced Research Projects Activity (IARPA) via Air Force Research Laboratory (AFRL) contract number FA8650-10-C-7058.

Please visit our poster ID : 02

Examples : Set Expansion

20

Examples : ASIA

21

Set Expansion

22

ASIA Task

23