| | SLO | ENG | Cookies and privacy

Bigger font | Smaller font

Show document

Title:Identifikacija pomanjkljive kode na osnovi mejnih vrednosti programskih metrik
Authors:Beranič, Tina (Author)
Heričko, Marjan (Mentor) More about this mentor... New window
Files:.pdf DOK_Beranic_Tina_2018.pdf (8,61 MB)
 
Language:Slovenian
Work type:Doctoral dissertation (mb31)
Typology:2.08 - Doctoral Dissertation
Organization:FERI - Faculty of Electrical Engineering and Computer Science
Abstract:Programske metrike so pomemben element procesov zagotavljanja in kontrole kakovosti programskih rešitev. Za njihovo učinkovito uporabo potrebujemo mejne vrednosti, s katerimi lahko med drugim zaznamo tudi pomanjkljivo kodo. Da identificiramo resnično pomanjkljive programske entitete, je potrebno mejne vrednosti programskih metrik določiti na reprezentativen in zanesljiv način. Na osnovi sistematičnega pregleda literature ter primerjave obstoječih pristopov določanja mejnih vrednosti smo izbrano metodo izračuna, ki temelji na statistični analizi in primerjalnih podatkih, nadgradili in uporabili nad 400 projekti v štirih objektnih programskih jezikih. Za devet izbranih metrik smo analizirali sistematično pridobljene mejne vrednosti ter ugotovili, da so odvisne od programskega jezika. Pridobljene mejne vrednosti so osnova v disertaciji predlaganega pristopa k identifikaciji pomanjkljive programske kode. Pristop temelji na kombinaciji programskih metrik, pri čemer je kakovost obravnavanih programskih entitet ovrednotena z uporabo funkcije večine. Z uporabo predlaganega pristopa se število potencialno pomanjkljivih entitet bistveno omeji, zmanjša pa se tudi število lažno pozitivnih rezultatov. Validacijo rezultatov identifikacije smo izvedli s pomočjo potrditvene študije, v sklopu katere je sodelovalo 43 ocenjevalcev, ki so z razvitim orodjem za sodelovanje in podporo presojam ovrednotili 131 entitet v treh različnih programskih jezikih. Strokovne presoje potrjujejo zanesljivost vrednotenja kakovosti razredov na osnovi predlaganega pristopa tako glede natančnosti kot točnosti izvedene identifikacije pomanjkljive programske kode.
Keywords:kakovost programske opreme, pomanjkljive programske entitete, metrike programske opreme, primerjalni podatki, porazdelitev metričnih vrednosti, primerjava mejnih vrednosti, strokovna presoja
Year of publishing:2018
Publisher:[T. Beranič]
Source:Maribor
UDC:004.4\'2:004.415.4(043.3)
COBISS_ID:22002198 Link is opened in a new window
NUK URN:URN:SI:UM:DK:AONI001D
License:CC BY-NC-ND 4.0
This work is available under this license: Creative Commons Attribution Non-Commercial No Derivatives 4.0 International
Views:471
Downloads:96
Metadata:XML RDF-CHPDL DC-XML DC-RDF
Categories:KTFMB - FERI
:
  
Average score:(0 votes)
Your score:Voting is allowed only for logged in users.
Share:AddThis
AddThis uses cookies that require your consent. Edit consent...

Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Secondary language

Language:English
Title:Identification of deficient code based on software metric thresholds
Abstract:An important part of software quality assurance is detecting and managing existing code smells within software. Code smells can be detected with exceeding threshold values of selected software metrics. To provide proper identification, it is important to calculate thresholds in a reliable and representative way, using an appropriate approach. Based on a systematic literature review and a comparison of existing approaches, we used an approach that depends on a statistical analysis of benchmark data, which resulted in different risk areas for nine selected software metrics. They were identified for four object-oriented programming languages and an analysis showed that threshold values are not language independent. Using the derived threshold values, code smell identification was carried out. To increase the reliability of the identification, we combined selected software metrics into five categories that evaluate different aspects of object-oriented software. The proposed approach is based on applying the majority function on a combination of software metrics. An empirical analysis has shown that the combination of categories and consequently, the consideration of more software metrics in the assessment process, can significantly reduce the number of potentially deficient entities. Also, the reliability of code smell detection is increased. To validate the identification of code smells, a study of expert judgment was performed which confirmed the correctness and reliability of the conducted identification.
Keywords:software quality, deficient program entities, software metrics, benchmark data, distribution of metric values, threshold comparison, expert judgment


Comments

Leave comment

You have to log in to leave a comment.

Comments (0)
0 - 0 / 0
 
There are no comments!

Back
Logos of partners University of Maribor University of Ljubljana University of Primorska University of Nova Gorica