Get Another Label? Improving Data Quality and Data Mining Using Multiple, Noisy Labelers

DMCA

Get Another Label? Improving Data Quality and Data Mining Using Multiple, Noisy Labelers

Cached

Download Links

by Victor S. Sheng , Foster Provost , Panagiotis G. Ipeirotis

Citations:

252 - 12 self

BibTeX

@MISC{Sheng_getanother,
    author = {Victor S. Sheng and Foster Provost and Panagiotis G. Ipeirotis},
    title = {Get Another Label? Improving Data Quality and Data Mining Using Multiple, Noisy Labelers},
    year = {}
}

OpenURL

Abstract

This paper addresses the repeated acquisition of labels for data items when the labeling is imperfect. We examine the improvement (or lack thereof) in data quality via repeated labeling, and focus especially on the improvement of training labels for supervised induction. With the outsourcing of small tasks becoming easier, for example via Rent-A-Coder or Amazon’s Mechanical Turk, it often is possible to obtain less-than-expert labeling at low cost. With low-cost labeling, preparing the unlabeled part of the data can become considerably more expensive than labeling. We present repeated-labeling strategies of increasing complexity, and show several main results. (i) Repeated-labeling can improve label quality and model quality, but not always. (ii) When labels are noisy, repeated labeling can be preferable to single labeling even in the traditional setting where labels are not particularly cheap. (iii) As soon as the cost of processing the unlabeled data is not free, even the simple strategy of labeling everything multiple times can give considerable advantage. (iv) Repeatedly labeling a carefully chosen set of points is generally preferable, and we present a robust technique that combines different notions of uncertainty to select data points for which quality should be improved. The bottom line: the results show clearly that when labeling is not perfect, selective acquisition of multiple labels is a strategy that data miners should have in their repertoire; for certain label-quality/cost regimes, the benefit is substantial.

Keyphrases

data mining using multiple improving data quality noisy labelers traditional setting considerable advantage robust technique different notion everything multiple time low-cost labeling certain label-quality cost regime label quality present repeated-labeling strategy unlabeled data model quality multiple label simple strategy supervised induction selective acquisition repeated acquisition several main result amazon mechanical turk small task data quality training label bottom line data point unlabeled part low cost data item less-than-expert labeling

DMCA