scieee Open visual document viewer

FM fact label

Horcas Aguilera, José Miguel; Galindo Duarte, José Ángel; Fuentes, Lidia; Benavides Cuevas, David Felipe

Abstract

FM Fact Label is a tool for visualizing the characterizations of feature models based on their metadata, structural measures, and analytical metrics. Although there are various metrics available to characterize feature models, there is no standard method to visualize and identify unique properties of feature models. Unlike existing tools, FM Fact Label provides a standalone web-based platform for configurable and interactive visualization, enabling export to various formats. This contribution is significant because it supports the Universal Variability Language (UVL) and enhances the UVL ecosystem by offering a common representation of the results of existing analysis tools.

Full text

Science o Compu e P og amming 240 (2025) 103214 A ailable online 26 Sep embe 2024 0167-6423/© 2024 The Au ho (s). Published by Else ie B.V. This is an open access a icle unde he CC BY license (h p://c ea i ecommons.o g/licenses/by/4.0/). Con en s lis s a ailable a ScienceDi ec Science o Compu e P og amming jou nal homepage: www.else ie .com/loca e/scico O iginal so wa e publica ion FM ac label Jose-Miguel Ho casa,∗, Jose A. Galindob, Lidia Fuen esa, Da id Bena idesb aITIS So wa e, Uni e sidad de Málaga, Andalucía Tech, Spain bI3US, Uni e sidad de Se illa, Spain A R T I C L E I N F O A B S T R A C T Keywo ds: Cha ac e iza ion Fea u e model Me ics So wa e p oduc line Va iabili y Visualiza ion FM Fac Label is a ool o isualizing he cha ac e iza ions o ea u e models based on hei me ada a, s uc u al measu es, and analy ical me ics. Al hough he e a e a ious me ics a ailable o cha ac e ize ea u e models, he e is no s anda d me hod o isualize and iden i y unique p ope ies o ea u e models. Unlike exis ing ools, FM Fac Label p o ides a s andalone web-based pla o m o configu able and in e ac i e isualiza ion, enabling expo o a ious o ma s. This con ibu ion is significan because i suppo s he Uni e sal Va iabili y Language (UVL) and enhances he UVL ecosys em by offe ing a common ep esen a ion o he esul s o exis ing analysis ools. Me ada a See Table 1. Table 1 Code me ada a. N . Code me ada a desc ip ion Please fill in his column C1 Cu en code e sion 1.0.0 C2 Pe manen link o code/ eposi o y used o his code e sion h ps://doi.o g/10.5281/zenodo.12794159 C3 Pe manen link o Rep oducible Capsule h ps:// m ac label.adaby on.uma.es/ C4 Legal Code License GNU Gene al Public License (GPL) C5 Code e sioning sys em used gi C6 So wa e code languages, ools, and se ices used Py hon, Ja aSc ip , HTML, CSS C7 Compila ion equi emen s, ope a ing en i onmen s and dependencies Py hon 3.9, Flama, D3.js C8 Link o de elope documen a ion/manual h ps:// m ac label.gi hub.io/ C9 Suppo email o ques ions [email p o ec ed] 1. Mo i a ion and significance •Fea u e models (FMs) a e he de ac o s anda d o modeling a iabili y in so wa e p oduc lines (SPLs), and he Uni e sal Va i- abili y Language (UVL) is gaining ac ion as a ex ual language o speci ying FMs. In p ac ice, when handling la ge-scale FMs, * Co esponding au ho . E-mail add ess: [email p o ec ed] (J.-M. Ho cas). h ps://doi.o g/10.1016/j.scico.2024.103214 Recei ed 29 Feb ua y 2024; Recei ed in e ised o m 31 July 2024; Accep ed 23 Sep embe 2024 Science o Compu e P og amming 240 (2025) 103214 2 J.-M. Ho cas, J.A. Galindo, L. Fuen es e al. Fig. 1. So wa e a chi ec u e FM Fac Label. hei ex ual ep esen a ion (in UVL) o hei isual ep esen a ion (as a ea u e diag am) offe limi ed assis ance in pinpoin ing he dis inguishing p ope ies ha make an FM unique. FM Fac Label con ibu es o he UVL ecosys em by p o iding a ool o effec i ely isualize he in o ma ion o an FM in e ms o i s me ada a, and i s s uc u al and analy ical me ics. •FM Fac Label offe s a common ep esen a ion o he cha ac e iza ion o an FM in he o m o a ac label. Adhe ing o he p inciples and bes design p ac ices om he isualiza ion field [1], he ool p o ides an in e ac i e and adap able isualiza ion, enabling use s o cus omize he p esen a ion o he FM p ope ies. The ool has been used o analysis and ca ego iza ion o FMs [2,3], and is cu en ly being in eg a ed wi hin UVLHub [4], a eposi o y o FMs. •FM Fac Label is a s andalone web-based applica ion whe e he use uploads an FM and a isualiza ion o i s cha ac e iza ion is au oma ically gene a ed which can be cus omized and expo ed o se e al o ma s. • Se e al collec ions o me ics o FMs a e a ailable in he li e a u e [5,6], bu hey ocus on assessing he e alua ion o FM main ainabili y, igno ing de ailed desc ip ion o an FM beyond quali y. FM Fac Label allows SPL p ac i ione s o be e choice hei FMs in diffe en scena ios (e.g., eaching pu poses, scalabili y e alua ion, model compa isons). While cu en SPL ools such as Fea u eIDE [7], Glencoe [8], SPLOT [9], and DyMMe [10] epo syn ac ic and seman ic s a is ics o FMs, hey lack a isualiza ion componen o efficien ly con ey he esul s ou side he ool o IDE. FM Fac Label delega es he au oma ed analysis o FMs o flama [11], and uses D3 [12], a da a-d i en app oach o isualiza ion based on web s anda ds. Mo eo e , FM Fac Label p o ides mo e in o ma ion, cu en ly offe ing 46 me ics, compa ed o o he SPL ools such as DyMMe (38 me ics), Glencoe (27 me ics), Fea u eIDE (16 me ics), and SPLOT (14 me ics). 2. So wa e desc ip ion FM Fac Label is a web-based applica ion accessible online, which cons uc s a cha ac e iza ion o a gi en FM and p oduces i s isualiza ion in he o m o a ac label inspi ed by he nu i ion ac label commonly seen in he ood indus y. This isualiza ion aims o clea ly dis inguish a ious ypes o in o ma ion con ained wi hin he FM cha ac e iza ion: (1) The name o he FM; (2) Me ada a o he FM such as a b ie desc ip ion, keywo ds, au ho (s), yea , domain, and DOI o he pape whe e i is published; (3) The s uc u al me ics o syn a ical s a is ics o he FM, including he ea u e ee ela ions, c oss- ee cons ain s, and o he pa ame e s; and (4) The analysis esul s o seman ical me ics ob ained by easoning on he FM (i.e., using a SAT o a BDD sol e ). 2.1. So wa e a chi ec u e Fig. 1ou lines he ool’s a chi ec u e and echnologies. The ool i sel offe s a web se ice o upload he FM and i s me ada a ia an online o m (Web Se ice componen ). I suppo s FMs in UVL and Fea u eIDE o ma s. The FM Cha ac e iza ion module in he se e -side ga he s and manages all he FM in o ma ion. We dis inguish h ee kinds o in o ma ion: me ada a (FM Me ada a), s uc u al me ics (FM Me ics), and analysis esul s (FM Analysis), ea ing all o hem as an FM p ope y. Each FM P ope y includes a name, a desc ip ion, and a pa en p ope y o hie a chical o ganiza ion in he ac label. P ope ies a e associa ed wi h an FM P ope y Measu e ha p o ides he specific alues o he p ope y. Fo ins ance, he lis o abs ac ea u es, hei size, and Science o Compu e P og amming 240 (2025) 103214 3 J.-M. Ho cas, J.A. Galindo, L. Fuen es e al. Fig. 2. Fac label o an FM cha ac e iza ion and main unc ionali ies o he ool. a io o he ABSTRACT FEATURES p ope y. Analysis asks a e delega ed o ex e nal ools, wi h he cu en implemen a ion elying on flama [11](da k componen in Fig. 1). The ool encompasses 46 measu es, wi h he possibili y o being ex ended o include addi ional me ics om he SPL li e a u e. All in o ma ion included in each FM P ope y and i s FM P ope y Measu e is se ialized in JSON o send i o he clien -side. The ac label isualiza ion is gene a ed using D3 [12]on he clien -side (Visualiza ion componen ). D3 is based on web s anda ds (HTML, CSS, Ja aSc ip , SVG, and JSON) o in eg a e isualiza ion componen s and employs a da a-d i en app oach o dynamic ans o ma ions. 2.2. So wa e unc ionali ies The gene a ed ac label is (1) comple ely cus omizable, and (2) in e ac i e. This means, fi s , ha all me ada a a e op ional, allowing use s o inpu only ele an in o ma ion. Second, configu able op ions enable use s o selec i ely collapse o expand FM sub-p ope ies o hide p ope ies wi h ze o o emp y alues. Addi ionally, ho e ing o e a p ope y e eals i s desc ip ion in a ool ip, while clicking on i e eals i s specific alue. Thi d, he configu ed isualiza ion can be expo ed in high-quali y SVG o ma o inclusion in esea ch publica ions o as a PNG image o online po abili y. Las ly, he en i e FM cha ac e iza ion can be expo ed in JSON o ma o plain ex o in e ope abili y wi h o he SPL ools o u he analysis such as s a is ical analysis o FMs’ da ase s. 3. Illus a i e examples Fig. 2depic s he ac label isualiza ion and majo unc ions o he ool. Ten eal ea u e models a e p o ided wi hin he ool as examples o i s usage which can also be downloaded. A ac label o he Linux ea u e model (one o he la ges models in he SPL communi y) is a ailable in [13]. 4. Impac •FM Fac Label imp o es he cha ac e iza ion o FMs beyond suppo ing hei quali y e alua ion. The ool allows SPL p ac i ione s o answe ques ions such as which FMs can be used o e alua e a new algo i hm implemen ed o op imize configu a ions?, o which a e he p ope ies o he FMs ha effec he scalabili y o you p oposal?, o which FM is mo e app op ia e o eaching pu poses?, among o he s. •By including all undamen al in o ma ion, no jus main enance me ics, he ool allows SPL p ac i ione s o be e selec hei FMs in diffe en scena ios. Fo ins ance, o eaching pu poses, a small FM ha includes a leas one ea u e and ela ion o each ype (e.g., an o -g oup, an xo -g oup, a equi es and an excludes cons ain , a dead- ea u e) is mo e app op ia e han a la ge Science o Compu e P og amming 240 (2025) 103214 4 J.-M. Ho cas, J.A. Galindo, L. Fuen es e al. FM ha canno e en be comple ely isualized in a ea u e diag am. In con as , o e alua e he scalabili y o a new analysis ope a ion, a la ge-scale FM wi h housands o ea u es and cons ain s may be p e e able [3]. • Thanks o FM Fac Label ool, use s can now ocus on he FM i sel wi hou he bu den o acqui ing o analyzing i s p ope ies. The e is no need o use s o ins all o manage sepa a e analysis ools. •F om he fi s elease o he FM Fac Label in 2022 [13], he ool has ecei ed 13 ci a ions acco ding o Google Schola , includ- ing [2–4]. •The ool is cu en ly being in eg a ed in o UVLHub [4]. UVLHub is a eposi o y o FMs da ase s using UVL and Open Science p inciples de eloped by se e al pa ne s o he SPL communi y. 5. Conclusions FM Fac Label acili a es he gene a ion o common isualiza ions ep esen ing FM cha ac e iza ions h ough he in eg a ion o FM me ada a, syn ac ical me ics, and analy ical da a. I s adap able a chi ec u e allows o seamless in eg a ion in o he UVL ecosys em and easy ex ension o accommoda e o he FM analysis ools and addi ional me ics. This ool helps SPL p ac i ione s o disce n he unique a ibu es o FMs and selec he mos sui able FM o hei specific equi emen s. We an icipa e he in eg a ion o FM Fac Label in o a ious ools as an independen componen [4]. 6. Fu u e plans We plan o expand FM Fac Label o cha ac e ize da ase s o FMs, and inco po a e addi ional ea u es such as h esholds o he me ics, and configu a ion op ions such as a ho izon al isualiza ion o he label. A comple e lis o planned ea u es is a ailable in he Gi Hub p ojec . CRediT au ho ship con ibu ion s a emen Jose-Miguel Ho cas: W i ing – o iginal d a , Visualiza ion, So wa e, Me hodology, In es iga ion, Concep ualiza ion. Jose A. Galindo: W i ing – e iew & edi ing, Visualiza ion, Valida ion, So wa e, Fo mal analysis. Lidia Fuen es: W i ing – e iew & edi ing, Valida ion, Supe ision, Resou ces, Funding acquisi ion, Concep ualiza ion. Da id Bena ides: W i ing – e iew & edi ing, Resou ces, P ojec adminis a ion, Me hodology, Funding acquisi ion, Concep ualiza ion. Decla a ion o compe ing in e es The au ho s decla e he ollowing financial in e es s/pe sonal ela ionships which may be conside ed as po en ial compe ing in e es s: Lidia Fuen es, Da id Bena ides epo s financial suppo and a el we e p o ided by Minis y o Science, Inno a ion and Uni e si ies (Spain). Lidia Fuen es, Da id Bena ides epo s financial suppo and a el we e p o ided by Jun a de Andalucía. Mónica Pin o, edi o o his special issue, also con ibu ed o he fi s e sion o his p ojec . I he e a e o he au ho s, hey decla e ha hey ha e no known compe ing financial in e es s o pe sonal ela ionships ha could ha e appea ed o influence he wo k epo ed in his pape . Acknowledgemen s We hank Mónica Pin o who has also con ibu ed o his p ojec . This wo k was pa ially suppo ed by EU, FEDER/Minis y o Sci- ence, Inno a ion and Uni e si ies/Jun a de Andalucía/S a e Resea ch Agency/CDTI wi h he ollowing g an s: TASOVA PLUS esea ch ne wo k (RED2022-134337-T), IRIS (PID2021-122812OB-I00) (FEDER unds), DAEMON (H2020-101017109), Da a-pl (PID2022- 138486OB-I00), AquaIA (GOPG-SE-23-0011) and MIDAS (IDI-20230256). Funding o open access cha ge: Uni e sidad de Málaga/ CBUA. Re e ences [1] D. Rome o-O gan idez, J.M. Ho cas, J.A. Galindo, D. Bena ides, Da a isualiza ion guidance using a so wa e p oduc line app oach, J. Sys . So w. 213 (2024) 112029, h ps://doi .o g /10 .1016 /J .JSS .2024 .112029. [2] J.A. Galindo, A.J. Dominguez, J. Whi e, D. Bena ides, La ge language models o gene a e meaning ul ea u e model ins ances, in: 27 h ACM In e na ional Sys ems and So wa e P oduc Line Con e ence (SPLC), ol. A, 2023, pp. 15–26. [3] D. Munoz, M. Pin o, L. Fuen es, D.S. Ba o y, T ans o ming nume ical ea u e models in o p oposi ional o mulas and he uni e sal a iabili y language, J. Sys . So w. 204 (2023) 111770, h ps://doi .o g /10 .1016 /J .JSS .2023 .111770. [4] D. Rome o-O gan idez, J.A. Galindo, C. Sunde mann, J.-M. Ho cas, D. Bena ides, UVLHub: a ea u e model da a eposi o y using u l and open science p inciples, J. Sys . So w. 216 (2024) 112150, h ps://doi .o g /10 .1016 /j .jss .2024 .112150. [5] E. Baghe i, D. Gase ic, Assessing he main ainabili y o so wa e p oduc line ea u e models using s uc u al me ics, So w. Qual. J. 19 (3) (2011) 579–612, h ps://doi .o g /10 .1007 /s11219 -010 -9127 -2. [6] C.I.M. Beze a, R.M.C. And ade, J.M. Mon ei o, Explo ing quali y measu es o he e alua ion o ea u e models: a case s udy, J. Sys . So w. 131 (2017) 366–385, h ps://doi .o g /10 .1016 /j .jss .2016 .07 .040. [7] T. Thüm, C. Käs ne , F. Benduhn, J. Meinicke, G. Saake, T. Leich, Fea u eIDE: an ex ensible amewo k o ea u e-o ien ed so wa e de elopmen , Sci. Compu . P og am. 79 (2014), h ps://doi .o g /10 .1016 /j .scico .2012 .06 .002. [8] A. Schmi , C. Be inge , G. Rock, Glencoe – a ool o specifica ion, isualiza ion and o mal analysis o p oduc lines, in: 25 h T ans, Eng, 2018, pp. 665–673. Science o Compu e P og amming 240 (2025) 103214 5 J.-M. Ho cas, J.A. Galindo, L. Fuen es e al. [9] M. Mendonca, M. B anco, D. Cowan, S.P.L.O.T.: so wa e p oduc lines online ools, in: OOPSLA, 2009. [10] C.I.M. Beze a, R. Lima, P. Sil a, Dymme 2.0: a ool o dynamic modeling and e alua ion o ea u e model, in: 35 h B azilian Symposium on So wa e Enginee ing (SBES), 2021, pp. 121–126. [11] J.A. Galindo, J.M. Ho cas, A. Fel e nig, D. Fe nández-Amo ós, D. Bena ides, FLAMA: a collabo a i e effo o build a new amewo k o he au oma ed analysis o ea u e models, in: 27 h ACM In e na ional Sys ems and So wa e P oduc Line Con e ence (SPLC), ol. B, 2023, pp. 16–19. [12] M. Bos ock, V. Ogie e sky, J. Hee , D3da a-d i en documen s, IEEE T ans. Vis. Compu . G aph. 17 (12) (2011), h ps://doi .o g /10 .1109 /TVCG .2011 .185. [13] J.M. Ho cas, J.A. Galindo, M. Pin o, L. Fuen es, D. Bena ides, FM Fac Label: a configu able and in e ac i e isualiza ion o ea u e model cha ac e iza ions, in: 26 h ACM In e na ional Sys ems and So wa e P oduc Line Con e ence (SPLC), ol. B, 2022, pp. 42–45.