ML-based data-entry automation and data anomaly detection to support data quality assurance

BELGACEM, Hichem

Download

Doctoral thesis (Dissertations and theses)

ML-based data-entry automation and data anomaly detection to support data quality assurance

BELGACEM, Hichem

2023

Permalink
https://hdl.handle.net/10993/57433

Files (1)Send to Details Statistics Bibliography Similar publications

Files

Full Text

Thesis_HB.pdf

Author postprint (1.77 MB)

Download

All documents in ORBilu are protected by a user license.

Send to

RIS BibTex APA Chicago Permalink X Linkedin

Details

Keywords :

Form filling, Data entry forms, Machine Learning, Software data quality, User interfaces

Abstract :

[en] Data playsacentralroleinmodernsoftwaresystems,whichare very oftenpoweredbymachinelearning(ML)andusedincriticaldo- mains ofourdailylives,suchasfinance,health,andtransportation. However,theeffectivenessofML-intensivesoftwareapplicationshighly depends onthequalityofthedata.Dataqualityisaffectedbydata anomalies; dataentryerrorsareoneofthemainsourcesofanomalies. The goalofthisthesisistodevelopapproachestoensuredataquality by preventingdataentryerrorsduringtheform-fillingprocessandby checking theofflinedatasavedindatabases. The maincontributionsofthisthesisare: 1. LAFF, anapproachtoautomaticallysuggestpossiblevaluesofcat- egorical fieldsindataentryforms. 2. LACQUER, anapproachtoautomaticallyrelaxthecompleteness requirementofdataentryformsbydecidingwhenafieldshould be optionalbasedonthefilledfieldsandhistoricalinputinstances. 3. LAFF-AD, anapproachtoautomaticallydetectdataanomaliesin categorical columnsinofflinedatasets. LAFF andLACQUERfocusmainlyonpreventingdataentryerrors during theform-fillingprocess.Bothapproachescanbeintegratedinto data entryapplicationsasefficientandeffectivestrategiestoassistthe user duringtheform-fillingprocess.LAFF-ADcanbeusedofflineon existing suspiciousdatatoeffectivelydetectanomaliesincategorical data. In addition,weperformedanextensiveevaluationofthethreeap- proaches,assessingtheireffectivenessandefficiency,usingreal-world datasets.

Disciplines :

Computer science

Author, co-author :

BELGACEM, Hichem ; University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > SVV

Language :

English

Title :

ML-based data-entry automation and data anomaly detection to support data quality assurance

Defense date :

15 September 2023

Institution :

Unilu - University of Luxembourg [The Faculty of Sciences, Technology and Medicine], Luxembourg, Luxembourg

Degree :

Docteur en Informatique (DIP_DOC_0006_B)

Promotor :

BIANCULLI, Domenico ; University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > SVV

President :

BRIAND, Lionel ; University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > SVV

Jury member :

Boytsov, Andrey

SHIN, Seung Yeob ; University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > SVV

Baresi, Luciano; Politecnico Di Milano

Available on ORBilu :

since 20 November 2023

Statistics

Number of views

150 (19 by Unilu)

Number of downloads

187 (5 by Unilu)

More statistics