App store mining for feature extraction: analyzing user reviews
Autor(a) principal: | |
---|---|
Data de Publicação: | 2023 |
Outros Autores: | , |
Tipo de documento: | Artigo |
Idioma: | eng |
Título da fonte: | Acta scientiarum. Technology (Online) |
Texto Completo: | http://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/article/view/62867 |
Resumo: | A recent study shows that the most commonly used app stores, iOS app store and Android's play store had up to 2 million apps from where users can explore, purchase, share, download and install applications on a single click. More and more apps are being added daily which makes the app stores a large software repository. An enormous amount of data is provided by the end users in the form of user reviews. This data can provide valuable insights for acquiring requirements. User reviews include plenty of information as they contain information about faulty features (Bug reports), ideas for new features and improvements (feature requests), or user experience that can help app developers and vendors to achieve software enhancement and evolution tasks. As feature requests are the ones that are most helpful for the purpose of eliciting new requirements, the work is done on feature requests out of the 3 categories mentioned above. This study is conducted to provide a general approach that extracts feature request from user reviews. The proposed approach has five main building blocks, namely, (i) Extraction of Feature Requests, (ii) Feature Extraction from Feature Requests, (iii) topic modelling, (iv) sentiment analysis and (v) Classification into Functional Requirements (FR) and Non-Functional Requirements (NFR). Firstly, it finds the feature requests out of the reviews, then perform extraction of features from feature requests, then further work on grouping the features into topics, next apply sentiment analysis to mine the user opinions on the extracted topic and finally group them into Functional and Non-Functional requirements. This article provides the app developers a more user-centred definition of requirements and improvements. At the topic modelling phase, the results received the highest coherence score, 0.70, with k=22 topics. Sentiment analysis is used to classify feature request, with an accuracy of 80.20%, precision of 84.25%, and recall of 80.42%. With an accuracy of 84.4%, requirements are classified quite successfully as well. |
id |
UEM-6_4bdb842382cfd3d30bad76b07e5dec23 |
---|---|
oai_identifier_str |
oai:periodicos.uem.br/ojs:article/62867 |
network_acronym_str |
UEM-6 |
network_name_str |
Acta scientiarum. Technology (Online) |
repository_id_str |
|
spelling |
App store mining for feature extraction: analyzing user reviewsApp store mining for feature extraction: analyzing user reviewsfeature enhancement; feature requests; requirement engineering; user reviews; non-negative matrix factorizationfeature enhancement; feature requests; requirement engineering; user reviews; non-negative matrix factorizationA recent study shows that the most commonly used app stores, iOS app store and Android's play store had up to 2 million apps from where users can explore, purchase, share, download and install applications on a single click. More and more apps are being added daily which makes the app stores a large software repository. An enormous amount of data is provided by the end users in the form of user reviews. This data can provide valuable insights for acquiring requirements. User reviews include plenty of information as they contain information about faulty features (Bug reports), ideas for new features and improvements (feature requests), or user experience that can help app developers and vendors to achieve software enhancement and evolution tasks. As feature requests are the ones that are most helpful for the purpose of eliciting new requirements, the work is done on feature requests out of the 3 categories mentioned above. This study is conducted to provide a general approach that extracts feature request from user reviews. The proposed approach has five main building blocks, namely, (i) Extraction of Feature Requests, (ii) Feature Extraction from Feature Requests, (iii) topic modelling, (iv) sentiment analysis and (v) Classification into Functional Requirements (FR) and Non-Functional Requirements (NFR). Firstly, it finds the feature requests out of the reviews, then perform extraction of features from feature requests, then further work on grouping the features into topics, next apply sentiment analysis to mine the user opinions on the extracted topic and finally group them into Functional and Non-Functional requirements. This article provides the app developers a more user-centred definition of requirements and improvements. At the topic modelling phase, the results received the highest coherence score, 0.70, with k=22 topics. Sentiment analysis is used to classify feature request, with an accuracy of 80.20%, precision of 84.25%, and recall of 80.42%. With an accuracy of 84.4%, requirements are classified quite successfully as well.A recent study shows that the most commonly used app stores, iOS app store and Android's play store had up to 2 million apps from where users can explore, purchase, share, download and install applications on a single click. More and more apps are being added daily which makes the app stores a large software repository. An enormous amount of data is provided by the end users in the form of user reviews. This data can provide valuable insights for acquiring requirements. User reviews include plenty of information as they contain information about faulty features (Bug reports), ideas for new features and improvements (feature requests), or user experience that can help app developers and vendors to achieve software enhancement and evolution tasks. As feature requests are the ones that are most helpful for the purpose of eliciting new requirements, the work is done on feature requests out of the 3 categories mentioned above. This study is conducted to provide a general approach that extracts feature request from user reviews. The proposed approach has five main building blocks, namely, (i) Extraction of Feature Requests, (ii) Feature Extraction from Feature Requests, (iii) topic modelling, (iv) sentiment analysis and (v) Classification into Functional Requirements (FR) and Non-Functional Requirements (NFR). Firstly, it finds the feature requests out of the reviews, then perform extraction of features from feature requests, then further work on grouping the features into topics, next apply sentiment analysis to mine the user opinions on the extracted topic and finally group them into Functional and Non-Functional requirements. This article provides the app developers a more user-centred definition of requirements and improvements. At the topic modelling phase, the results received the highest coherence score, 0.70, with k=22 topics. Sentiment analysis is used to classify feature request, with an accuracy of 80.20%, precision of 84.25%, and recall of 80.42%. With an accuracy of 84.4%, requirements are classified quite successfully as well.Universidade Estadual De Maringá2023-11-06info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionapplication/pdfhttp://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/article/view/6286710.4025/actascitechnol.v46i1.62867Acta Scientiarum. Technology; Vol 46 No 1 (2024): Em proceso; e62867Acta Scientiarum. Technology; v. 46 n. 1 (2024): Publicação contínua; e628671806-25631807-8664reponame:Acta scientiarum. Technology (Online)instname:Universidade Estadual de Maringá (UEM)instacron:UEMenghttp://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/article/view/62867/751375156635Copyright (c) 2024 Acta Scientiarum. Technologyhttp://creativecommons.org/licenses/by/4.0info:eu-repo/semantics/openAccessMemon, Zulfiqar AliMunawar, NidaKamal, Maha2024-02-08T19:23:27Zoai:periodicos.uem.br/ojs:article/62867Revistahttps://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/indexPUBhttps://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/oai||actatech@uem.br1807-86641806-2563opendoar:2024-02-08T19:23:27Acta scientiarum. Technology (Online) - Universidade Estadual de Maringá (UEM)false |
dc.title.none.fl_str_mv |
App store mining for feature extraction: analyzing user reviews App store mining for feature extraction: analyzing user reviews |
title |
App store mining for feature extraction: analyzing user reviews |
spellingShingle |
App store mining for feature extraction: analyzing user reviews Memon, Zulfiqar Ali feature enhancement; feature requests; requirement engineering; user reviews; non-negative matrix factorization feature enhancement; feature requests; requirement engineering; user reviews; non-negative matrix factorization |
title_short |
App store mining for feature extraction: analyzing user reviews |
title_full |
App store mining for feature extraction: analyzing user reviews |
title_fullStr |
App store mining for feature extraction: analyzing user reviews |
title_full_unstemmed |
App store mining for feature extraction: analyzing user reviews |
title_sort |
App store mining for feature extraction: analyzing user reviews |
author |
Memon, Zulfiqar Ali |
author_facet |
Memon, Zulfiqar Ali Munawar, Nida Kamal, Maha |
author_role |
author |
author2 |
Munawar, Nida Kamal, Maha |
author2_role |
author author |
dc.contributor.author.fl_str_mv |
Memon, Zulfiqar Ali Munawar, Nida Kamal, Maha |
dc.subject.por.fl_str_mv |
feature enhancement; feature requests; requirement engineering; user reviews; non-negative matrix factorization feature enhancement; feature requests; requirement engineering; user reviews; non-negative matrix factorization |
topic |
feature enhancement; feature requests; requirement engineering; user reviews; non-negative matrix factorization feature enhancement; feature requests; requirement engineering; user reviews; non-negative matrix factorization |
description |
A recent study shows that the most commonly used app stores, iOS app store and Android's play store had up to 2 million apps from where users can explore, purchase, share, download and install applications on a single click. More and more apps are being added daily which makes the app stores a large software repository. An enormous amount of data is provided by the end users in the form of user reviews. This data can provide valuable insights for acquiring requirements. User reviews include plenty of information as they contain information about faulty features (Bug reports), ideas for new features and improvements (feature requests), or user experience that can help app developers and vendors to achieve software enhancement and evolution tasks. As feature requests are the ones that are most helpful for the purpose of eliciting new requirements, the work is done on feature requests out of the 3 categories mentioned above. This study is conducted to provide a general approach that extracts feature request from user reviews. The proposed approach has five main building blocks, namely, (i) Extraction of Feature Requests, (ii) Feature Extraction from Feature Requests, (iii) topic modelling, (iv) sentiment analysis and (v) Classification into Functional Requirements (FR) and Non-Functional Requirements (NFR). Firstly, it finds the feature requests out of the reviews, then perform extraction of features from feature requests, then further work on grouping the features into topics, next apply sentiment analysis to mine the user opinions on the extracted topic and finally group them into Functional and Non-Functional requirements. This article provides the app developers a more user-centred definition of requirements and improvements. At the topic modelling phase, the results received the highest coherence score, 0.70, with k=22 topics. Sentiment analysis is used to classify feature request, with an accuracy of 80.20%, precision of 84.25%, and recall of 80.42%. With an accuracy of 84.4%, requirements are classified quite successfully as well. |
publishDate |
2023 |
dc.date.none.fl_str_mv |
2023-11-06 |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/article info:eu-repo/semantics/publishedVersion |
format |
article |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/article/view/62867 10.4025/actascitechnol.v46i1.62867 |
url |
http://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/article/view/62867 |
identifier_str_mv |
10.4025/actascitechnol.v46i1.62867 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.relation.none.fl_str_mv |
http://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/article/view/62867/751375156635 |
dc.rights.driver.fl_str_mv |
Copyright (c) 2024 Acta Scientiarum. Technology http://creativecommons.org/licenses/by/4.0 info:eu-repo/semantics/openAccess |
rights_invalid_str_mv |
Copyright (c) 2024 Acta Scientiarum. Technology http://creativecommons.org/licenses/by/4.0 |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
application/pdf |
dc.publisher.none.fl_str_mv |
Universidade Estadual De Maringá |
publisher.none.fl_str_mv |
Universidade Estadual De Maringá |
dc.source.none.fl_str_mv |
Acta Scientiarum. Technology; Vol 46 No 1 (2024): Em proceso; e62867 Acta Scientiarum. Technology; v. 46 n. 1 (2024): Publicação contínua; e62867 1806-2563 1807-8664 reponame:Acta scientiarum. Technology (Online) instname:Universidade Estadual de Maringá (UEM) instacron:UEM |
instname_str |
Universidade Estadual de Maringá (UEM) |
instacron_str |
UEM |
institution |
UEM |
reponame_str |
Acta scientiarum. Technology (Online) |
collection |
Acta scientiarum. Technology (Online) |
repository.name.fl_str_mv |
Acta scientiarum. Technology (Online) - Universidade Estadual de Maringá (UEM) |
repository.mail.fl_str_mv |
||actatech@uem.br |
_version_ |
1799315338145300480 |