ptype: probabilistic type inference

Taha Ceritli; Christopher K. I. Williams; James Geddes

doi:10.1007/s10618-020-00680-1

ptype: probabilistic type inference

Taha Ceritli, Christopher K. I. Williams, James Geddes

Research output: Contribution to journal › Article › peer-review

Abstract

Type inference refers to the task of inferring the data type of a given column of data. Current approaches often fail when data contains missing data and anomalies, which are found commonly in real-world data sets. In this paper, we propose ptype, a probabilistic robust type inference method that allows us to detect such entries, and infer data types. We further show that the proposed method outperforms existing methods.

Original language	English
Pages (from-to)	870-904
Number of pages	35
Journal	Data Mining and Knowledge Discovery
Volume	34
Issue number	3
Early online date	16 Mar 2020
DOIs	https://doi.org/10.1007/s10618-020-00680-1
Publication status	Published - 31 May 2020

Keywords / Materials (for Non-textual outputs)

Type inference
Robustness
Probabilistic finite-state machine

Access to Document

10.1007/s10618-020-00680-1Licence: Creative Commons: Attribution (CC-BY)

ptype_CERITLI_DOA24022020_AFVAccepted author manuscript, 2.32 MBLicence: Creative Commons: Attribution (CC-BY)
ptype_CERITLI_DOA24022020_VOR_CC-BYFinal published version, 1.17 MBLicence: Creative Commons: Attribution (CC-BY)

https://link.springer.com/article/10.1007/s10618-020-00680-1Licence: Creative Commons: Attribution (CC-BY)

Cite this

@article{29aa1b5d10504ca2b0ec071ebfabf2a8,

title = "ptype: probabilistic type inference",

abstract = "Type inference refers to the task of inferring the data type of a given column of data. Current approaches often fail when data contains missing data and anomalies, which are found commonly in real-world data sets. In this paper, we propose ptype, a probabilistic robust type inference method that allows us to detect such entries, and infer data types. We further show that the proposed method outperforms existing methods.",

keywords = "Type inference, Robustness, Probabilistic finite-state machine",

author = "Taha Ceritli and Williams, \{Christopher K. I.\} and James Geddes",

year = "2020",

month = may,

day = "31",

doi = "10.1007/s10618-020-00680-1",

language = "English",

volume = "34",

pages = "870--904",

journal = "Data Mining and Knowledge Discovery",

issn = "1573-756X",

publisher = "Springer",

number = "3",

}

TY - JOUR

T1 - ptype: probabilistic type inference

AU - Ceritli, Taha

AU - Williams, Christopher K. I.

AU - Geddes, James

PY - 2020/5/31

Y1 - 2020/5/31

N2 - Type inference refers to the task of inferring the data type of a given column of data. Current approaches often fail when data contains missing data and anomalies, which are found commonly in real-world data sets. In this paper, we propose ptype, a probabilistic robust type inference method that allows us to detect such entries, and infer data types. We further show that the proposed method outperforms existing methods.

AB - Type inference refers to the task of inferring the data type of a given column of data. Current approaches often fail when data contains missing data and anomalies, which are found commonly in real-world data sets. In this paper, we propose ptype, a probabilistic robust type inference method that allows us to detect such entries, and infer data types. We further show that the proposed method outperforms existing methods.

KW - Type inference

KW - Robustness

KW - Probabilistic finite-state machine

U2 - 10.1007/s10618-020-00680-1

DO - 10.1007/s10618-020-00680-1

M3 - Article

SN - 1573-756X

VL - 34

SP - 870

EP - 904

JO - Data Mining and Knowledge Discovery

JF - Data Mining and Knowledge Discovery

IS - 3

ER -

ptype: probabilistic type inference

Abstract

Keywords / Materials (for Non-textual outputs)

Access to Document

Fingerprint

Cite this