Endre søk
RefereraExporteraLink to record
Permanent link

Direct link
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
Exploration on Automated Software Requirement Document Readability Approaches
Blekinge Tekniska Högskola, Fakulteten för datavetenskaper, Institutionen för programvaruteknik.
Blekinge Tekniska Högskola, Fakulteten för datavetenskaper, Institutionen för programvaruteknik.
2017 (engelsk)Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
Abstract [en]

Context. The requirements analysis phase, as the very beginning of software development process, has been identified as a quite important phase in the software development lifecycle. Software Requirement Specification (SRS) is the output of requirements analysis phase, whose quality factors play an important role in the evaluation work. Readability is a quite important SRS quality factor, but there are few available automated approaches for readability measurement, because of the tight dependency on readers' perceptions. Low readability of SRS documents has a serious impact on the whole process of software development. Therefore, it's extremely urgent to propose effective automated approaches for SRS documents readability measurement. Using traditional readability indexes to analyze readability of SRS documents automatically is a potentially feasible approach. However, the effectiveness of this approach is not systematically evaluated before.

Objectives. In this study, firstly, we aim to understand the readability of texts and investigate approaches to score texts readability manually. Then investigate existing automated readability approaches for texts with their working theories. Next, evaluate the effectiveness of measuring the readability of SRS documents by using these automated readability approaches. Finally, rank these automated approaches by their effectiveness.

Methods. In order to find out the way how human score the readability of texts manually and investigate existing automated readability approaches for texts, systematic literature review is chosen as the research methodology. Experiment is chosen to explore the effectiveness of automated readability approaches.

Results. We find 67 articles after performing systematic literature review. According to systematic literature review, human judging the readability of texts through reading is the most common way of scoring texts readability manually. Additionally, we find four available automated readability assessments tools and seven available automated readability assessments formulas. After executing the experiment, we find the actual value of effectiveness of all selected approaches are not high and Coh-Metrix presents the highest actual value of effectiveness of automated readability approach among the selected approaches.

Conclusions. Coh-Metrix is the most effective automated readability approach, but the feasibility in directly applying Coh-Metrix in SRS documents readability assessments cannot be permitted. Since the actual value of evaluated effectiveness is not high enough. In addition, all selected approaches are based on metrics of readability measures, but no semantic factors are blended in readability assessments. Hence studying more on human perception quantifying and adding semantic analysis in SRS documents readability assessment could be two research directions in future.

sted, utgiver, år, opplag, sider
2017.
Emneord [en]
Readability Measurement, Software Requirement Specification, Automated Approach
HSV kategori
Identifikatorer
URN: urn:nbn:se:bth-14816OAI: oai:DiVA.org:bth-14816DiVA, id: diva2:1118454
Fag / kurs
PA2534 Master's Thesis (120 credits) in Software Engineering
Utdanningsprogram
PAAXA Master of Science Programme in Software Engineering
Veileder
Tilgjengelig fra: 2017-07-03 Laget: 2017-06-30 Sist oppdatert: 2018-01-13bibliografisk kontrollert

Open Access i DiVA

fulltext(2439 kB)264 nedlastinger
Filinformasjon
Fil FULLTEXT02.pdfFilstørrelse 2439 kBChecksum SHA-512
95cf626fc75ae8fffe02133c3ba4420cce29c4d8abc2da72d1246760fc31244af6dbc67c6bd9b44fc3f0287831e975721c65ff6d36519f3afa6671a8d9a7c66d
Type fulltextMimetype application/pdf

Søk i DiVA

Av forfatter/redaktør
Chen, MingdaHe, Yao
Av organisasjonen

Søk utenfor DiVA

GoogleGoogle Scholar
Totalt: 264 nedlastinger
Antall nedlastinger er summen av alle nedlastinger av alle fulltekster. Det kan for eksempel være tidligere versjoner som er ikke lenger tilgjengelige

urn-nbn

Altmetric

urn-nbn
Totalt: 452 treff
RefereraExporteraLink to record
Permanent link

Direct link
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf