Name
Abstract | ||
|---|---|---|
Due to the unreliable environment in mobile could, attribute values or tuples may be missing or lost. Thus we should capture missing values to make data mining and analysis more accurate. Besides ignoring or setting to default values, many imputation methods have been proposed, but they also have their limitations. This paper proposes a human-machine hybrid workflow to study the missing value filling method with crowdsourcing. First we propose a missing value selection algorithm to select the missing values which are suitable to use crowdsourcing for filling. Then we propose three missing values filling methods according to different attribute types to select answers from crowdsourcing. Experimental results show that our algorithms could improve data quality significantly with low costs. |
Year | DOI | Venue |
|---|---|---|
2014 | 10.1007/978-3-319-07782-6_70 | WASA |
Keywords | Field | DocType |
data cleaning, missing values, crowdsourcing | Data mining,Data quality,Tuple,Crowdsourcing,Computer science,Selection algorithm,Missing data,Imputation (statistics),Workflow | Conference |
Volume | ISSN | Citations |
8491 | 0302-9743 | 3 |
PageRank | References | Authors |
0.38 | 14 | 2 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Chen Ye | 1 | 8 | 4.16 |
| Hongzhi Wang | 2 | 421 | 73.72 |