Sign inSign up
Open-domain Quantity Queries on Web Tables: Annotation, Response, and Consensus Models thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Open-domain Quantity Queries on Web Tables: Annotation, Response, and Consensus Models

Published on Oct 07, 20141773 Views

Sunita Sarawagi

Over 40% of columns in hundreds of millions of Web tables contain numeric quantities. Tables are a richer source of structured knowledge than free text. We harness Web tables to answer queries whose t

Research Sessions

Related categories

Data MiningKnowledge Extraction

Chapter list

escape velocity jupiter00:00
Web search00:04
Web tables00:15
To bridge the gap00:39
Step 1: Matching query to table03:12
Step 2: Number and Unit extraction - 104:36
Step 2: Number and Unit extraction - 205:00
Step 2: Number and Unit extraction - 305:41
Collective inference - 107:40
Collective inference - 208:07
Collective inference - 308:27
Collective inference - 410:05
Collective inference - 510:25
Unit Annotation10:51
Initial annotator: set of rules….did not work11:33
Our approach12:23
An example parse13:15
An example parse14:12
Benefits of collective evaluation14:20
Accuracy of QuantCFG14:51
Summary15:03
Thank you.15:34