In e na ional Jou nal o Compu a ional In elligence Sys ems
Vol. 12(2), 2019, pp. 661–675
DOI: h ps://doi.o g/10.2991/ijcis.d.190613.001; ISSN: 1875-6891; eISSN: 1875-6883
h ps://www.a lan is-p ess.com/jou nals/ijcis/
A Compu e -Based Suppo Sys em o Coope a i e Tasks
in Nu sing Homes
Juan M. Albe ola1,3, Elena del Val4, Angelo Cos a2, Paulo No ais2, Vicen e Julián1,*
1Depa amen o de Sis emas In o má icos y Compu ación (DSIC), Uni e si a Poli ècnica de València, València, Spain
2ALGORITMI Cen e , Uni e si y o Minho, B aga, Po ugal
3Flo ida Uni e si à ia, Ca a oja, Spain
4Uni e sidad de Za agoza, Te uel, Spain
ARTICLE INFO
A icle His o y
Recei ed 12 No 2018
Accep ed 13 May 2019
Keywo ds
Compu e -based suppo ool
G oup o ma ion
IT Applica ions in Heal h Ca e
Coali ion s uc u e
Coope a i e g oups
A B S T R A C T
Di e en s udies ha e shown he bene i s o a coope a i e ac i i ies p og amme o he elde ly. Membe s o a g oup wi h sim-
ila abili ies o disabili ies a e o en encou aged by ha ing he oppo uni y o sha e hei expe iences, knowledge, o opinions.
Ne e heless, when ca egi e s y o plan speci ic coope a i e ac i i ies, di e en aspec s, as he indi idual needs o each pe son,
should be aken in o accoun , which no ably inc eases he complexi y o ha plani ica ion. This pape p oposes a compu e -
based suppo ool o ec ea ional he apis s which acili a es he managemen ask o g ouping elde ly people in o coope a i e
g oups o exis ing ac i i ies. To do his, an i e a i e lea ning p ocess is p oposed allowing he o ma ion o p ope dis ibu ions
o elde ly people in o ac i i ies.
©2019 The Au ho s. Published by A lan is P ess SARL.
This is an open access a icle dis ibu ed unde he CC BY-NC 4.0 license (h p://c ea i ecommons.o g/licenses/by-nc/4.0/).
1. INTRODUCTION
G oup wo k is a widely used modali y o wo k among olde peo-
ple o achie e di e en objec i es such as (i) socializa ion he apy,
educa ion, and ec ea ion, (ii) se ice and ad ocacy, (iii) suppo ,
(i ) he apy, and ( ) amily ca e and ca e-gi e assis ance. This ype
o wo k g ea ly imp o es he social in e ac ion and adap a ion o
he elde ly. Coope a i e ac i i ies play a signi ican ole in nu sing
homes as hey help elde ly home esiden s o adap o a new way o
li e. These ac i i ies a e essen ial o p og essi ely achie e he apeu-
ic, educa ional, o social goals in se ings such as hospi als, elde -
ca e cen e s, nu sing homes, and social o ganiza ions.
One o he esul s ob ained h ough pa icipa ion in coope a i e
ac i i ies is he achie emen o a sense o accomplishmen and he
oppo uni y ha elde ly home esiden s ha e o de elop hei skills,
which p o ides hem wi h an imp o emen in sel -es eem and sa -
is ac ion [1]. Nu sing homes and mo e speci ically ca egi e s ha
wo k in hese ins i u ions ha e o ca y ou an ac i i y planning
whe e he cha ac e is ics o he po en ial pa icipan s (i.e., he esi-
den s) should be conside ed. These cha ac e is ics a e he physical
abili ies, social ela ions, an eceden s, and in e es s o he esiden s.
In gene al, in mos scena ios, he social in e ac ion ha is gene -
a ed in coope a i e ac i i ies is mo e impo an han he ac i i y
i sel [1].
*Co esponding au ho . Email: [email protected] .es
Ne e heless, only a ew wo ks ha e ied o analyze which is he
bes way o o ganize olde people in o ac i i y g oups. Typically,
g oups a e andomly o med o o ganized by he ca egi e s ha use
hei common sense. Membe s o a g oup wi h simila abili ies a e
o en encou aged o sha e hei li e expe ience, knowledge, and gen-
e al iews and opinions wi h hei pee s and ca e s.
The ec ea ional he apis has a g ea challenge in planning ac i i-
ies o esiden s. This challenge consis s on dis ibu ing esiden s
in he di e en ac i i ies in he mos app op ia e way (i.e., phys-
ical and cogni i e capaci ies o he esiden s co espond o hose
equi ed in he ac i i ies) [2–4]. In o de o each he expec ed goals
h ough coope a i e ac i i ies, ec ea ional he apis s also ha e o
wo k oge he wi h he esidence nu sing s a . Nu sing s a helps
he ec ea ional he apis o selec he ac i i y ha will be mos ben-
e icial o he needs o each indi idual esiden .
1.1. Ques ionnai e
This wo k was alida ed by wo ke s o dayca e cen e s in he no h-
e n a ea o Po ugal, like he Cen o Social I mandade de S. To ca o.
The alida ion was pe o med h ough a ques ionnai e ha was
done o he ca egi e s ( egis e ed nu ses and medical pe sonnel),
iden i ying hei needs and expec ed esul s o helping ools. Two
ques ionnai es we e esponded by 12 pe sons, one be o e he ool
was p esen ed o hem, being used as a equi emen s analysis; and
he o he ques ionnai e was pos -ope a ion o he ool, p o iding a
Pd _Folio:661
662 J.M. Albe ola e al. / In e na ional Jou nal o Compu a ional In elligence Sys ems 12(2) 661–675
alida ion o he ope a ion. The ques ionnai e was designed simple
and using mos o he ime Like scales o ep esen he esponses
o he ques ioned people. This educes he sp eading o he possible
esponses and linea izes he a ained alues.
In Table 1, we con i m ha mos o he dayca e cen e s con ac ed
ha e some kind o compu e so wa e ha acks he esiden s in o -
ma ion. The ype o in o ma ion ha i is a ailable is unknown, bu
a he leas he e is a basic knowledge abou who a e he clien s o he
dayca e cen e . Table 1also shows ha all he ques ionnai e pa ici-
pan s know how o ope a e a compu e , hus a echnological ba ie
does no exis .
In Figu e 1, he opinions a e dis ibu ed be ween a social se ice
echnician and a social assis an , while he es o opinions we e
a ibu ed o he amily o o he s, meaning ha he wo ke s eel ha
he amily should ha e a say in he planning o he ac i i ies. In his
speci ic en i onmen , he social se ice echnician has a ank abo e
he social assis an and has mo e expe ise abou he ac i i ies and
hei impac o he use s. The e o e, i is implied ha mos o he
ques ionnai e pa icipan s belie e ha he ask o scheduling ac i -
i ies should be pe o med by a highly quali ied pe son. Fou people
ha e chosen no o pa icipa e on his ques ion.
In he esponse o Figu e 2 he esponden s ha e educed he impo -
ance o ac i i ies planning, pu ing his ask as only mode a ely
impo an .
Figu e 3p esen s he amoun o collabo a ion he esiden s do on
he ac i i y planning. This means ha he pe son ha is in cha ge o
he planning knows ha he opinion o he people ha pe o m he
ac i i ies ma e , and ha ing hem choosing he nex ac i i y esul s
Table 1 Usage o compu e sys ems.
Yes (%) No (%)
Is he e an in o ma ion sys em whe e
esiden s’ da a and ac i i ies a e s o ed?
91.7 8.3
Do you eel com o able wo king wi h
compu e s?
100 0
Figu e 1 In you opinion, wha is he p o essional p o ile ha he
pe son esponsible o ca ying ou planning ac i i ies should ha e?
Figu e 2 How would you classi y he s a emen : “Is ac i i y
planning a big p oblem a esidencies?”
on he inc ease o happiness le el. The e o e, he pe sonali y and
likes o he esiden s play an impo an pa in ac i i ies planning.
Al hough he ques ionnai e esponden s ha e a i med ha he
planning o ac i i ies do no ha e a c i ical impo ance, hey ehe-
men ly a i m ha hey would like a ool ha helps hem in ac i i-
ies planning, showed in Figu e 4.
In Table 2 he esponden s ha e showed a clea endency o spend
la ge amoun s o ime (o e 1 hou ) dedica ed o planning ac i i ies,
hus i is clea ha i is a ele an ask on he dayca e ope a ion.
In his ques ionnai e he pe sonal opinion o he ca egi e s was also
measu ed. This measu es p o ides a be e insigh o he planning
ask and how he ca egi e eels abou i . Fu he mo e, he pe sonal
expe ience is ga he ed, being i al o unde s and wha a e he ele-
an esiden ’s ea u es o ake in o accoun in he ool.
In Table 3 he esponden s di ide hei opinion in h ee opics, being
one common c i ical issue is he economic means a ailable. I is
clea ha wi hou unding he dayca e cen e s a e limi ed o asks
ha hey ha e he means o o ganize, hus he e is a high le el o
ac i i ies epe i ion.
In Figu e 5i is clea ha he ac i i ies a e limi ed by he budge
bu o pe o m ac i i ies he e is no he need o ha e a la ge bud-
ge . Like s a ed be o e, he lack o esou ces limi he a ie y o he
ac i i ies. This ac o may lead he esiden s o momen s o bo e-
dom and incompliance.
Figu e 3 How much do esiden s collabo a e in he selec ion o
ac i i ies?
Figu e 4 Do you ind i in e es ing o ha e a ool o help you
plan ac i i ies?
Table 2 How much ime akes o plan a
g oup ac i i y, accoun ing he esiden s’
in e es s and limi a ions?
<1 hou 25%
1 o 4 hou s 60%
4 o 8 hou s 0%
>8 hou s 15%
Pd _Folio:662
J.M. Albe ola e al. / In e na ional Jou nal o Compu a ional In elligence Sys ems 12(2) 661–675 663
Table 3 Wha ac o s no ela ed o esiden s a e aken in o accoun
when planning ac i i ies?
S a , ime, and means 71.6%
Economic a ailabili y and ac i i ies ha go
agains he p e e ence o all pa icipan s
14.2%
Economic a ailabili y, accessibili y,
a ailabili y, and swimming ac i i ies
14.2%
Figu e 5 Do you hink mone a y es ain s s ongly in luence
planning ac i i ies?
In Table 4, he esponden s do no belie e ha a he e ogeneous
g oup o people is impo an o an ac i i y, no hey a i m
ha simila people should be bundled oge he . Addi ionally, hey
a i m ha he esiden s should be inqui ed abou how hey el
abou he pe o med ac i i y, being he ca egi e s able o make
an in o med decision abou wha ac i i ies hey would like o
pe o m nex .
This ques ionnai e alida es ou ini ial idea abou he necessi y
o imp o e he o ganiza ion o olde people in o ac i i y g oups,
and he esponses highligh ea u es ha should be used on he
algo i hms. The social and wo k expe ience o he ques ionees a e
in aluable, and wi h his knowledge we a e able o build a decision
suppo sys em ha helps he ca egi e s o plan ac i i ies and make
g oups o people ha a e happy o pe o m hem.
This pape builds on a compu a ional eamwo k heo y o iden-
i y oppo uni ies o echnology o suppo he c ea ion and coo -
dina ion o g oups o ac i i ies ha enable and enhance exis ing
app oaches. Speci ically, he pape p oposes a ool o ec ea ional
he apis s and ca egi e s which acili a es he managemen ask o
g ouping elde ly people in o op imal g oups o exis ing ac i i ies.
The p oposed ool elies on he p o ile o he elde ly people, coali-
ional s uc u e gene a ion, and an i e a i e lea ning p ocess which
allows o ming p ope dis ibu ions o elde ly people in o ac i i-
ies. This ool should be used as a suppo sys em, being he esul s
supe ised by he ec ea ional he apis s and ca egi e s, hus alle-
ia ing he e o and ime ha hey dedica e o ac i i ies planning.
The es o he pape is s uc u ed as ollows: Sec ion 2p esen s
di e en p e ious wo ks ha y o analyze he bene i s o olde
adul s wi h ac i e pa icipa ion in di e en ypes o ac i i ies and
he impo ance o co ec ly iden i y g oups ha can pe o m spe-
ci ic ac i i ies. Sec ion 3analyses in de ail he p oblem ha we y
o sol e. Sec ion 4shows he p oposed model o g oup gene a ion.
Sec ion 5p esen s a case s udy. Sec ion 6desc ibes he expe imen s
and he p oposal alida ion, and, inally, in Sec ion 7some conclu-
sions a e exposed.
2. RELATED WORK
As s a ed abo e, he e a e some s udies ha show he sociologi-
cal connec ion be ween ac i e li e p omo ing ac i i ies and elde ly
people. In [5] a s udy is p esen ed analyzing how well-being can be
enhanced modi ying exe cise classes and inc easing pa icipa ion.
Mo eo e , in [6] he impo ance o physical exe cise in olde people
is s udied. The au ho s ied o p edic he le el o physical ac i i y
equi ed o ensu e ha a pe sons physical s a us was in he ange
be ween clinically a o able o heal hy.
O he s udies we e cen e ed in a speci ic ype o ac i i ies, like
leisu e ac i i ies. Leisu e pa icipa ion has been p o en o be ben-
e icial and has a posi i e link o success ul ageing, gene a ing a
g ea e in ol emen in leisu e ac i i ies. This in ol emen was asso-
cia ed wi h be e heal h in olde age [7]. The wo k p esen ed in
[8] showed how he p omo ion o ac i e pa icipa ion in di e si-
ied leisu e ac i i ies should also be emphasized in iew o he pos-
sibili y o g ea e bene i s gained om i s in ol emen . Mo eo e ,
[9] ema ks he impo ance o de eloping communi y-based p o-
g ams o acili a e physical ac i i ies o olde people wi h a ch onic
diseases.p
Undoub edly, he in luence o engagemen in men al, physical,
social, p oduc i e, and ec ea ional ac i i ies e ains, and in some
cases educes, he demen ia incidence in olde people [2]. Ene ge ic
ac i i ies ha in ol e men al o psychosocial componen s may ac
as s imuli o p ese e cogni ion [2].
Ano he impo an issue is he use o social ac i i ies by ca e-
gi e s. In o mal social ac i i y has accumula ed he mos e idence
o an in luence on well-being. In his sense, some s udies ha e ana-
lyzed he impo ance o he g oup o imp o e well-being. As an
example, he s udy p esen ed in [3] in es iga ed he capaci y o
g oup decision-making o build a sense o sha ed social iden i y
among elde ly ca e home esiden s and he eby inc ease hei social
engagemen , well-being, and cogni i e pe o mance. In he same
line, in [4] i was es ed how nu sing home popula ions engagemen
was highes among esiden s wi h adequa e unc ioning in ac i i ies
o daily li ing (ADL) and cogni ion.
The e a e se e al so wa e and ools, like AccuPoin Med and
Ca eVoyan , ha a e wi hin he elde ly ca e ealm bu a e only
Table 4 Planning con igu a ion opinion.
Ve y Impo an
(%)
Impo an (%) Mode a ly
Impo an (%)
Sligh ly
Impo an (%)
No Impo an
(%)
Do you hink i is impo an o mix people
wi h di e en cogni i e o physical
p oblems in hei ac i i ies?
16.7 16.7 41.7 8.3 16.7
A e he pa icipan s’ opinions abou each
ac i i y aken in o accoun when planning
he nex ac i i ies?
41.7 41.7 8.3 8.3 0
Pd _Folio:663
664 J.M. Albe ola e al. / In e na ional Jou nal o Compu a ional In elligence Sys ems 12(2) 661–675
designed o manage he ins i u ions and he s a , hus, no o man-
age he ac i i ies o ime o he ca e- ecei e s. The e o e, hey lack
he essen ial ea u es ha could enable he p o ision o he ac i e
aging pa adigm. The e a e also o he p ojec s di ec ed o he elde ly
and hei ca egi e s ha all unde he Ambien Assis ed Li ing
(AAL) domain such as [10–13]. AAL p ojec s p oduce sys ems and
p oduc s ha help i s’ use s o become independen and p o ides
aid on ADL’s, and a he same ime, c ea e a secu e and sa e en i-
onmen . The ocus is o imp o e he quali y o li e and s abilize he
heal h condi ion.
Mo eo e , we can ind speci ic so wa e and se ices which a e
di ec ed o p o essional ca egi e s, bu concen a e on p i a e com-
panies. This is he case o Home eam, Tabula P o so wa e, and
Sagely. These pla o ms show ha he e is a la ge in es men in
e ms o solu ions o moni o and ca e o elde ly people, whe he
hey a e a home o a esidencies. Mos o hem a e almos pu ely
adminis a i e, and easy he bu eauc a ic p ocess ha is keeping he
pe sonal heal h eco d upda ed as well as keeping ack o he ac i -
i ies and gene al well-being o he ca e- ecei e s. F om ou poin
o iew, i is clea ha hese applica ions equi e ha he ca egi e s
spend a la ge amoun o ime in e ac ing wi h hem and do no
ha e any au oma ed decision suppo sys em ha helps hem when
planning ac i i ies o de ec any decay o he ca e- ecei e s heal h
condi ion.
Howe e , he e a e no e y many s udies ha ocus on he key issue
o how o c ea e coope a i e g oups o elde ly, which is a ask ha
equi es a conside able amoun o ime o compose well-balanced
g oups (i.e., how o dis ibu e elde ly people in o ac i i ies in a way
ha all he people in each ac i i y sa is y he equi ed condi ions
o do i ). Mo e speci ically, when he e is a la ge numbe o olde
people and di e en g ouping c i e ia, he ask o o ming eams o
p omo e success ul ou pu s (i.e., well-being, engagemen and
in ol emen in he g oup) is conside ed an exponen ial p oblem.
Mo eo e , ac i i ies in g oups o eams can b ing posi i e and nega-
i e consequences (i.e., posi i e: social engagemen , well-being and
nega i e: lack o in e es , lack o mo i a ion, ense social con ex ),
making he ask o eam o ma ion a complex one. Se e al ac o s
like pe sonali y, heal h, and human beha io can in e e e wi h he
pe o mance o he g oup du ing an ac i i y [14]. The e o e, i is o
c ucial impo ance o iden i y g oups ha can pe o m co ec ly an
ac i i y.
The e a e a eas such as human esou ces managemen and collab-
o a i e lea ning whe e he p oblem o gene a ing g oups acco ding
o a se o p ede ined c i e ia (i.e., knowledge and expe ience, pe -
sonali y, lea ning/ hinking s yle) has been ho oughly s udied. In
hese a eas, compu a ional p oposals ha e been de eloped in o de
o au oma e he p ocess o g oup o ma ion in cases whe e i was
no manually iable. An example o au oma ion o he p ocess o
g oup o ma ion is p esen ed by Lin e al. [15]. The au ho s p esen
an algo i hm ha akes in o accoun wo c i e ia o he c ea ion
o lea ning g oups: he le els o unde s anding and he in e es s o
he s uden s. Ano he p oposal ha uses a compu a ional model
is p esen ed by Ca anaugh e al. [16]. The au ho s conside c i e-
ia such as gende , skills, and schedules o s uden s o au oma ic
gene a ion o g oups h ough a web ool. Wi e al. [17] p esen
a p oposal based on gene ic algo i hms whose goal is he gene -
a ion o g oups in R&D-o ien ed ins i u es. The algo i hm p o-
posed uses a uzzy model o collec in o ma ion abou candida es’
knowledge and expe ience on ce ain opics ela ed o a p ojec .
Ano he aspec conside ed by he algo i hm is he social ne wo k o
he candida es in o de o es ablish hei sui abili y o p ojec man-
agemen posi ions.
Taking in o accoun all o hese, as a as we know he e is no wo k
which ies o help ca egi e s o analyze how o g oup olde peo-
ple in o de o enhance ac i i y engagemen and well-being. Ou
p oposal p esen s a decision suppo ool ha gene a es a mo e
accu a e dis ibu ion o ca e- ecei e s o e ac i i ies. In o de o
gene a e he g oups o each ac i i y, he ool akes in o accoun ca e-
ecei e s’ physical s a e, ac i i y p e e ences, social ela ionships,
and p e ious ac i i ies al eady pe o med. A e each ac i i y, ca e-
ecei e s p o ide eedback abou he ac i i y (i.e., hei p e e ences)
and he ela ionships wi h o he membe s o he g oup. The ac i i y
is also added o hei his o ical ac i i y p o ile o a oid mono ony
o ac i i ies. As he numbe o ac i i ies pe o med inc eases (i.e.,
he e is mo e in o ma ion a ailable abou ca e- ecei e s), he ool
is able o lea n om his in o ma ion o imp o e he dis ibu ion o
ac i i ies ha be e i s wi h ca e- ecei e s’ expec a ions.
3. PROBLEM DESCRIPTION
Nu sing homes, and ins i u ions ha ca e o elde ly people, ha e
di icul ies in c ea ing an ha monious en i onmen ha can p o-
ide he equi ed a en ion ha hei ca e- ecei e s demand. Mos
o hese places y o p o ide an ac i e and heal hy li es yle by p o-
mo ing di e en coope a i e ac i i ies ha s imula e hem in e ms
o he physical and psychological aspec s. These ac i i ies can be
simple asks such as eading and able games o ad anced asks such
as dancing and hea e enac men , designed o engage wi h hem
and soo h hei likes.
The mos common issues highligh ed by he ca e- ecei e s a e he
lack o in e es ing and di e en ac i i ies and hei gene al disin-
e es [3,4]. These issues o igina e om ou elemen s: he lack o
ca egi e s [18], high numbe o physically and/o psychologically
challenged ca e- ecei e s [3,4], lack o e en s [19], and poo ac i -
i ies managemen [20].
In e ms o pe sonnel, mos o he Eu opean coun ies ha e es ab-
lished ules ha de ine he minimal pe sonnel ha ce i ied ins i-
u ions mus ha e [21], while some jus es ablish a social app oach
like “ha ing a numbe o people a ibu ed o each nu se ha is no
o e whelming”. Fo ins ance, Po ugal es ablishes he ollowing pe -
sonnel o esiden homes (whe e esiden s li e pe manen ly) [22]:
•A pa - ime social assis an (o ec ea ional he apis );
•One egis e ed nu se and an assis an o e e y 40
ca e- ecei e s;
•One nu sing assis an o e e y 8 ca e- ecei e s;
•A di ec ca e nu se o e e y 20 elde ly and nigh su eillance.
Es ablishmen s ha ca e o highly ulne able senio s equi e he
ollowing pe sonnel a io:
•A ull- ime social assis an (o ec ea ional he apis );
•A egis e ed nu se o e e y 20 elde ly;
Pd _Folio:664
J.M. Albe ola e al. / In e na ional Jou nal o Compu a ional In elligence Sys ems 12(2) 661–675 665
•A nu sing assis an o e e y 5 senio s;
•A di ec ca e nu se o e e y 15 elde ly.
These pe sonnel equi emen s a e he ba e minimum possible o
main ain a ha monious en i onmen while moni o ing he ca e-
ecei e s. The issue wi h his en i onmen is ha i lea es no space
o indi idual ca e and he numbe o ac i i ies ha can be pe -
o med become limi ed o ones ha a e include a la ge amoun o
people and ha a e easy o be moni o ed.
The lack o e en s and app op ia e planning means ha he ins i u-
ion does no p o ide plans ha a e indica ed o he ca e- ecei e s
and/o lack unding o hos complex ac i i ies ha equi e addi-
ional unding, o example, museum isi s.
Some o he ins i u ions ha e a ec ea ional he apis ha manages
he global and pe sonal asks ha c ea es ac i i ies plans acco ding
o he ins i u ion limi a ions (as s a ed be o e) and he use s likes.
Fu he mo e, he ac i i ies a e designed o be inclusi e, meaning
ha hey can se e a la ge amoun o people, being he issue ha a
la ge pa o hese ac i i ies may no be o he in e es o he ca e-
ecei e s, hus being emo ionally nega i e o hem. Meanwhile, he
ole o he ec ea ional he apis is c ucial as i coo dina es wi h
he egis e ed nu ses, nu sing assis an s, and di ec ca e nu ses he
immedia e needs and he likes o he use s, bu mos o he ime
he ac i i ies a e no op imized due o se e al social and mone a y
p oblems.
In ecen s udies [23,24] wo issues we e ou lined: he medical and
he o ganiza ional. Bo h equi e di e en le els o decision ha may
collide, acco ding o he ele ance and impac on he ca e- ecei e s
and on he ins i u ion. Focusing on he o ganiza ional issue, he e
a e h ee le els o ope a ion: he s a egic le el, he ac ical le el,
and he ope a ional le el. Each o hem has a g ea impac on he
ins i u ions ope a ions, being he ope a ional le el ans e sal o
he ca e- ecei e s, since i is ela ed o he human esou ces and ca e
ac i i ies scheduling p oblems, and daily in en o y managemen .
They e eal ha he numbe o cons ain s ( ela i e o he numbe
o human esou ces and he numbe o ac i i ies) di ec ly a ec s
he complexi y o achie ing an op imized solu ion (o i i is e en
achie able). [5] shows ha o main ain an ac i e li e, he elde ly
should pe o m ac i i ies ha concen a e in hei cogni ion, physi-
cal, and social li e. Also, ca e- ecei e s ha e esponded posi i ely o
he expe iences in he es s made in [5], and ha hey had imp o ed
he elde ly emo ional s a e e en i hey do no ha e ac i ely pa ic-
ipa ed in hose ac i i ies.
Bu he e is an issue in e ms o o ganizing a ious ac i i ies. The
Uni ed Na ions [25] ha e es ablished ha elde ly people (65 yea s
old and o e ) ha e medical, physical, and social equi emen s ha
a e no easy o be me . The e o e, mos o he ac i i ies ha day-ca e
cen e s do a e sa e and ha e a low physical impac . The issue is ha
he day-ca e esiden s equen ly epo high le els o dissa is ac ion
abou ac i i ies epe i i eness [3,4].
Accoun ing o all equi emen s o he ca e- ecei e s, law, en i on-
men al, and mone a y is complex and equi es esea ching all he
possible combina ions ha comply wi h he ha d and so equi e-
men s. Fo ins ance, i an ins i u ion wan s o o ganize a ip o a
museum i has o accoun o he ca e- ecei e s ha will be able o
go (acco ding o hei psychological and physical abili ies), i he e
is enough s a o ca e o hem, how can he ca e- ecei e s can be
pai ed (acco ding o hei social connec ion), wha ac i i ies can
be pe o med, among o he s equi emen s. Thus, he unde lying
di icul y unde mines he p omo ion o hese e en s and hei exe-
cu ion. The use o compu a ional sys ems ha may acili a e his
p ocess may p o ide he help ha he ins i u ions equi e o gi e
hei ca e- ecei e s joy ul ac i i ies and p omo e an ac i e li e and
ha monious en i onmen s.
These e idences we e ound on he esea ch p ocess o he Cogni-
i e Li e Assis an (CLA) [26,27]. The CLA is a cogni i e assis an
pla o m inse ed in an AAL en i onmen . The co e aim is o p o-
ide assis ance in ADL’s by p o iding in o ma ion abou he daily
e en s ha he use s ha e o pe o m, connec ing wi h ela i es and
iends and p omo ing he ac i e aging ini ia i e. I also suppo s
embedding ex e nal modules, such as use s emo ion de ec ion [28]
and pe suasion ecommenda ion [29].
On he de elopmen o ools o aid he ca egi e s ( ela i es, assis an
nu ses, egis e ed nu ses, o p i a e se ices companies) keep ack
o he e en s, whe eabou s, and heal h s a e o he ca e- ecei e s i
was ound ha he e was a se e e lack o o he solu ions o hem.
The ones ha exis ed we e managing ools designed o keep ack
o pe sonnel and accoun ing. One ou s anding complain om he
ca egi e s wi h mul iple ca e- ecei e s assigned was he lack o
g oups managemen . The e o e, g oup managing and mul iple peo-
ple e en s became one o he aims o he CLA.
The objec i e is o c ea e a solu ion ha can be easily in eg a ed
wi h he CLA and ha b idges he ask o g oup c ea ion and e en
choosing wi h he ca egi e s, elie ing he s ess o he ca egi e s.
Mo eo e , i will inc ease he ange o he CLA aid o he ca e-
ecei e s by p omo ing ac i i ies ha hey enjoy and wi h people
ha hey like. The ollowing sec ion shows he model needed o
achie e his.
4. PROPOSED MODEL
In his sec ion, we desc ibe ou policy o di iding ca e- ecei e s
in o g oups o pe o m ac i i ies each day du ing a pe iod o ime.
This policy elies on he p o iles o ca e- ecei e s (i.e., p e e ences,
heal h, iendship), he p o iles o ac i i ies, and he coali ion s uc-
u e gene a ion. Fi s , we desc ibe how di iding ca e- ecei e s in o
op imal g oup is equi alen o a Coali ion S uc u e Gene a ion
P oblem.
The Coali ion S uc u e Gene a ion p oblem e e s o pa i ioning
he componen s o a se in o exhaus i e and disjoin coali ions so
ha he global bene i s o he sys em a e op imized. In ou p oblem,
he componen s o he se a e he ca e- ecei e s ha ake pa in
coope a i e ac i i ies p oposed by a senio esidence cen e .
De ini ion 1. Le E={e1,...,en}be a se o ca e- ecei e s whe e
each indi idual eiis desc ibed by se o ea u es Fi ha desc ibes
his/he p o ile.
De ini ion 2. Le A={a1,...,am}be a se o ac i i ies planned o
a pe iod o ime (i.e., se e al weeks o mon hs depending on he
senio esidence cen e ). Each ac i i y is de ined by a se o equi e-
men s o being ca ied ou .
We de ine Gj∈Eas a subse o Ecalled g oup o indi iduals ha
pa icipa e in he same ac i i y. Conside ing a g oup Gj, i s alue
is gi en by a cha ac e is ic unc ion (Gj)∶ 2E→ ℝ ha assigns a
Pd _Folio:665

666 J.M. Albe ola e al. / In e na ional Jou nal o Compu a ional In elligence Sys ems 12(2) 661–675
eal- alued payo o Gj. The alue o a g oup (Gj)is calcula ed as
a linea combina ion o p unc ions 𝜙 ha calcula e di e en ypes
o ma ching depending on di e en inpu pa ame e s (e.g., he ea-
u es o each indi idual, he speci ic ac i i y):
Func ion: 𝜙i
Inpu : p1,…,pl: lis o pa ame e s
Ou pu : : eal [0.. 1]
And he alue o a g oup (Gj)is calcula ed as ollows:
(Gj)=∑
ei∈Gj
𝛼1⋅ 𝜙1+ 𝛼2⋅ 𝜙2…+𝛼p⋅ 𝜙p. (1)
whe e each 𝛼i ep esen s a weigh associa ed o he unc ion 𝜙iin
o de o gi e mo e o less ele ance o his unc ion o he calcula-
ion o he alue o he g oup. Gi en all hese weigh s, he ollowing
p ope y mus be ul illed:
p
∑
i=1
𝛼i= 1.
De ini ion 3. Ag oup s uc u e S ={G1,G2,...,Gk}is a pa i ion o
g oups such ha ∀i,j(i≠j),Gj∩Gi= ∅, ∪
∀Gj∈SGj=E. The alue
o a g oup s uc u e is deno ed by (S), whe e (S)is an e alua ion
unc ion o he g oup s uc u e. In his wo k, we conside ha he
alue o each g oup is independen o o he g oups. The e o e, we
can calcula e he alue o he g oup s uc u e as (S)=∑
Gj∈S
(Gj).
The goal o he algo i hm ha sol es ou p oblem is o de e -
mine an op imal g oup s uc u e o he o ganiza ion o ac i i ies
a gmax
S∈2E
(S).
I u ns ou ha pa i ioning a se o elde ly indi iduals in o dis-
join g oups while op imizing a social wel a e unc ion co esponds
o he o maliza ion o coali ion s uc u e gene a ion p oblems. In
o de o sol e his p oblem, we o mally de ine he coali ion s uc-
u e gene a ion p oblem as a linea p og amming p oblem [30] and
sol e i wi h he comme cial so wa e ILOG CPLEX 12.5.
4.1. Wo k low
The p oposed model o di iding ca e- ecei e s in o g oups o pe -
o m ac i i ies based on coali ion o ma ion s a egy is he basis o
a so wa e applica ion. The main goal o his applica ion is o acil-
i a e he managemen ask o g ouping ca e- ecei e s in o op imal
g oups o ac i i ies. Du ing he planning o he ac i i ies o se -
e al mon hs in elde ca e cen e s, a ec ea ional he apis may ca y
ou se e al ac i i ies ha equi e he o ma ion o g oups. In his
sec ion, we desc ibe he gene al wo k low and he mos impo an
ea u es o he p oposed ool o gene a e coali ions o each ac i i y.
As a gene al ou line (Figu e 6), he applica ion elies on he p o ile
o he elde ly pe son, coali ion s uc u e gene a ion, and an i e a i e
lea ning p ocess o o m p ope assignmen s o elde ly people in o
ac i i ies in a pe iod o ime. In he ollowing pa ag aphs, we explain
how hese elemen s a e pu oge he o p o ide an adequa e web ool
o acili a ing he planning o se e al mon hs o ac i i ies o elde ly
(see Figu e 7).
The ool has been de eloped o be in eg a ed in web pla o ms,
whe e ac o s (i.e., ec ea ional he apis s and ca e- ecei e s) can
in e ac wi h he sys em. The s a ing poin o he applica ion co -
esponds wi h he ec ea ional he apis (o o he membe o he
nu sing s a ) ha will egis e all he ca e- ecei e s and hei phys-
ical p o iles in he sys em (1). Once all he po en ial pa icipan s
ha e been egis e ed in he sys em, he ec ea ional he apis c ea es
he a ailable ac i i ies du ing he planned season (i.e., his pe iod
could be se e al weeks o mon hs depending on he a ailabili y o
esou ces o he ac i i ies). The ec ea ional he apis ills ou all
he in o ma ion abou he ac i i ies (e.g., ac i i y desc ip ion, du a-
ion, physical equi emen s). Mo eo e , he ec ea ional he apis
should de e mine he numbe o ac i i ies and he numbe o peo-
ple pe ac i i y. No e, ha hese ini ial asks could be a oided i
he elde ca e cen e al eady has an In o ma ion Sys em ha s o es
in o ma ion abou esiden s and ac i i ies.
Conside ing all o his in o ma ion, he g oup o ma ion mech-
anism is eady o s a he au oma ic gene a ion o g oups o
ac i i ies o each day o he planned season. I i is he i s ime
Figu e 6 Wo k low o he decision suppo ool.
Pd _Folio:666
J.M. Albe ola e al. / In e na ional Jou nal o Compu a ional In elligence Sys ems 12(2) 661–675 667
Figu e 7 Main page o he web applica ion o ac i i y g oup
gene a ion.
ha he ool akes as inpu a g oup o ca e- ecei e s, he e is no
eedback abou p e ious expe iences o he indi iduals (e.g., iend-
ships, p e e ences abou ac i i ies, his o ical ac i i y p o ile). The e-
o e, he ool only conside s he p o ile o he ca e- ecei e s, he
cha ac e is ics o he ac i i ies, he numbe o ac i i ies, and he
numbe o days o he planned season as inpu o he algo i hm
ha gene a es he g oups o he ac i i ies. O he wise, he algo i hm
can also conside in o ma ion p o ided by ca e- ecei e s as eed-
back om p e ious ac i i ies. As mo e ac i i ies a e pe o med by
he same popula ion, mo e in o ma ion will be a ailable abou hei
eedback. No e ha he ca egi e has he possibili y o modi ying
he weigh s o each o he ea u es conside ed o gene a e a sui able
alloca ion o indi iduals in ac i i ies. This ac allows he model o
adap o di e en scena ios. Then, he algo i hm will p o ide a sui -
able alloca ion o indi iduals in ac i i ies which will imp o e he
o e all sa is ac ion o he g oups (2). A his s age, he ca egi e has
he possibili y o modi y he sugges ed alloca ion o indi iduals in
ac i i ies acco ding o his/he p o essional c i e ia and expe ience
(3). Once he g oups a e c ea ed o pe o m an ac i i y, each indi-
idual is in o med (4).
When an ac i i y inishes, each pa icipan has he possibili y o log
in he applica ion in o de o p o ide eedback h ough a ques ion-
nai e. This ques ionnai e collec s in o ma ion abou his/he sa is-
ac ion ela ed o he ac i i y (e.g., i he/she conside s app op ia e
o upda e his/he p e e ences abou ac i i ies, and i he/she wan s
o upda e his/he ela ionship wi h he g oup membe s ha he/she
has in e ac ed wi h du ing he las ac i i y) (5). All his in o ma-
ion is s o ed in a da abase (6). I he pa icipan canno log in he
applica ion, he ca egi e in cha ge o he ac i i y can collec his
in o ma ion o ill he ques ionnai e. No e ha his in o ma ion is
op ional, he sys em does no equi e his in o ma ion o pe o m
he ac i i y planning, bu i p o ides mo e in o ma ion o a mo e
accu a e solu ion.
As new planning o ac i i ies is ca ied ou , new in o ma ion is ga h-
e ed and accumula ed so ha his in o ma ion can be used in u u e
g oup ac i i y assignmen s.
5. CASE STUDY
In his sec ion we p esen a Case S udy om a eal da ase . The es
subjec s we e c ea ed h ough he use o eal da a collec ed by a
Po uguese ins i u ion called Fundao Manuel F ancisco Clé igo and
p esen ed on hei annual epo (o 2015). Following, we desc ibe
how he da a was ins an ia ed by conside ing he model desc ibed
in Sec ion 4.
5.1. Speci ica ion
The da a ob ained was anonymized o he p o ec ion o he pa ic-
ipan s, hus keeping he esul s unbiased. F om he comple e in o -
ma ion abou he se e al housing and se ices ha he ins i u ion
has we ha e ocused on he pe manen esiden s. This decision was
done based on he di ec impac and he capabili y o comple e
moni o ing o he pa icipan s in u u e es s wi h eal subjec s.
Fu he mo e, hese subjec s a e he ones ha a e mo e exposed o
in e ac ions wi h he o he esiden s, as he mos o he ac i i ies
equi e ha hey pa icipa e. The epo is cons i u ed by he ol-
lowing in o ma ion:
•The gene al dis ibu ion and cha ac e is ics o he popula ion,
like he ollowing:
– Age and gende and hei dis ibu ion;
– Dis ibu ion o he diseases (wi hou ela ion o gende );
– Dis ibu ion o ac i i ies p e e ences;
– Physical esou ces and pe sonnel esou ces;
– Dependence le el o he pa icipan s (independen , pa ially
dependen , and ully dependen );
– Dis ibu ion o iendship ela ionships;
– Educa ion le el and o me p o ession.
•Planned ac i i ies acco ding o he dependence le el;
•Planned ac i i ies by mon h (one yea ).
•Execu ed ac i i ies mon hly (wi h da a o Janua y o Ap il).
Due o being anonymous, he in o ma ion was de ached om he
pa icipan s, hus we we e unable o co ela e he diseases, gende ,
age, and so on, wi h he speci ic pa icipan s. To ha e a da ase o
es subjec s a p ocedu e o co ela ion was implemen ed. This p o-
cedu e consis ed in aking he a ailable da a and co ela e andomly
he da a a ailable and gene a e pe sonas (a gene ic p ojec ion o
people). In e ms o associa ing diseases wi h he gene a ed pe -
sonas i was done by a andom p ocess wi h wo cons ains: depen-
dence le el and gende . Fo ins ance, i a speci ic use is classi ied
as independen no limi ing disease (like Alzheime ) is a ibu ed,
he same wi h mas ec omy ha was a ibu ed o he emale gende
(al hough i is possible o a male o su e his disease i is highly
imp obable, as mas ec omy epo s indica e). The p ocedu e was
pe o med in he ollowing way:
1. Name and gende co ela ion acco ding o he numbe o he
pa icipan s (24 emales and 19 males);
2. Co ela ion o he p e iously gene a ed pe sonas wi h he
dependence le els (5 independen , 17 pa ially dependen , and
21 ully dependen );
3. Clus e ing and co ela ion wi h diseases, wi h a mean o h ee
diseases pe es subjec and in acco dance o he p e iously
men ioned cons ain s;
Pd _Folio:667
668 J.M. Albe ola e al. / In e na ional Jou nal o Compu a ional In elligence Sys ems 12(2) 661–675
4. Associa ion be ween he pe sonas. F iendship le els ha a e
likes, dislikes, and indi e en , hus ollowing a uzzy logic pa -
e n. The iendship connec ion holds he symme ic ela ion;
5. Mapping o he ac i i ies pe o med each mon h and hei
dependence le eling.
The ou come o his p ocess is a da ase ep esen a i e o he possi-
ble pa ien s o a nu sing home. Al hough he andomiza ion can be
deba able, he in o ma ion p o ided equi ed some kind o associ-
a ion p ocess and i is ou opinion ha he mos unbiased way o
do i was o gene a e pe sonas ha we e gene a ed by a cons ained
andom p ocess.
In o de o de e mine he ac o s aken in o conside a ion when
assigning a pe son o an ac i i y, we asked he wo ke s o dayca e
cen e es abou his issue. As i can be obse ed in Table 5, he likes,
he physical, and he psychological condi ions a e he ac o s ha
a e he mos conside ed when he pa icipan s a e chosen o do an
ac i i y (i.e., hey a e conside ed as e y impo an by o e he 60%
o he ca egi e s). In second come o he ac o s such as he ela-
ionship wi h he esiden s, he ac i i y con ol, he equency, he
pe sonali y, and i needs a ca e ake .
These esponses may p o ide a bias o he algo i hm in he deci-
sion suppo sys em, as hey es ablish he p io i ies and wha needs
o be conside ed i s . Responses like he sphinc e con ol and spe-
cial medica ion may p o e ha some conside a ion is gi en o hese
issues bu a e no so ele an , al hough we do no know i he ac i i-
ies al eady con empla e hese issues, o i he esiden s o ca egi e s
a e p epa ed o hese issues’ con ingencies.
Fo he sake o simplici y, we selec only some o he mos ele-
an ac o s in o de o de ine he p o ile o each indi idual o he
expe imen s. We selec hose ac o s desc ibed as “ e y impo an ”
by o e he 60% o he ca egi e s (numbe s 2, 3, and 5), and hose
ac o s desc ibed as “ e y impo an ” o “impo an ” by o e he
80% o he ca egi e s (6 and 9). No e ha he model p oposed in
Sec ion 4would allow o add o he ac o s i needed. These ac o s
a e g ouped in o he ollowing ea u es:
•Physical and psychological s a us e e s o he physical and
psychological condi ions o he indi idual and can ake h ee
alues based on his/he medical p o ile: independen , pa ially
independen , o dependen . Depending on his s a us, he e a e
some ac i i ies ha a e mos sui able o an indi idual. This
ac o is assumed o be known om he beginning and emains
almos cons an du ing his/he s ay in he cen e .
•P e e ence o ac i i ies e e o he i s choices o a pe son
ela ed o an ac i i y, ha is, how appealing is an ac i i y o an
indi idual. This ea u e can ake h ee alues: appealing,
neu al, o nonappealing. We assume ha an indi idual has no
p e e ences un il he/she pa icipa es in a planned ac i i y. A e
his/he pa icipa ion, he elde ly has an opinion abou his
ac i i y and p o ides eedback abou his/he p e e ence.
•F iendship ela ionships ep esen he social ne wo k o he
senio esidence cen e . Nodes ep esen he indi iduals and
links a e conside ed as weigh ed bidi ec ional ela ions
be ween indi iduals and can ake h ee alues: non iends (i.e.,
he indi iduals conside each o he annoying), neu al (i.e., he
indi iduals a e indi e en wi h each o he ), o iends (i.e., he
indi iduals a e iends). Ini ially, in o ma ion o iendship
ela ionships is no a ailable. A e each ac i i y, indi iduals
p o ide in o ma ion abou his/he ela ionships wi h o he
ac i i y membe s.
•His o ical ac i i y p o ile s o es he sequence o ac i i ies
al eady pe o med by he indi idual du ing he planned
pe iod. This in o ma ion is conside ed in o de o a oid he
epe i ion o ac i i ies du ing a speci ic pe iod o ime.
A e he pa icipa ion in each ac i i y, new indi idual p e e ences
and iendship ela ionships a e lea ned and his o ical ac i i y p o-
iles a e modi ied. All o his in o ma ion is conside ed in u u e
g oup ac i i y con igu a ions (see Figu e 8).
The equi emen s associa ed o an ac i i y a e he ype and he
numbe o pa icipan s (see Figu e 9). We conside wo ypes o
ac i i ies based on hei equi emen s o pe o m hem: psycho-
logical and physical. Psychological ac i i ies include able games,
a is ic exp ession, eading, o eligious pa y among o he s. Physi-
cal ac i i ies include dancing, walks, ga dening, o culina y lessons
among o he s. The numbe o pa icipan s is ep esen ed as a ange
o nume ical alues ha de ines he minimum and maximum o
indi iduals equi ed o an ac i i y.
Acco ding o he ea u es ha desc ibe he p o ile o each indi id-
ual, we de ine he ollownig ou 𝜙 unc ions o calcula e he alue
o a g oup:
Table 5 Responses o “Wha ac o s do you conside impo an o assign a pe son o an ac i i y?”
Ve y Impo an
(%)
Impo an
(%)
Mode a ely
Impo an (%)
Sligh ly
Impo an (%)
No Impo an
(%)
1. The age 33.3 25 33.3 8.3 0
2. Physical condi ion 66.7 8.3 25 0 0
3. Psychological condi ion 66.7 8.3 16.7 8.3 0
4. The pe sonali y 16.7 58.3 16.7 8.3 0
5. Likes he ac i i y 75 16.7 8.3 0 0
6. Ra ely pe o ms he ac i i y 8.3 75 16.7 0 0
7. I he pa icipan s a e in con ol o i 25 50 16.7 8.3 0
8. I s cha ac e in ela ion o he ac i i y
(pa icipa i e, passi e, e c.)
8.3 41.7 50 0 0
9. Rela ionship wi h he esiden s 16.7 66.7 8.3 8.3 0
10. Possess sphinc e con ol 0 50 50 0 0
11. Is agg essi e 16.7 41.7 33.3 8.3 0
12. Is dominan o submissi e 8.3 41.7 50 0 0
13. Needs special medica ion 16.7 25 33.3 25 0
14. Needs a special ca e ake 16.7 50 25 8.3 0
Pd _Folio:668
J.M. Albe ola e al. / In e na ional Jou nal o Compu a ional In elligence Sys ems 12(2) 661–675 669
Figu e 8 Residen iew o he eedback ques ionnai e.
Figu e 9 Rec ea ional he apis iew o he ac i i y g oup gene a ion.
•Func ion 𝜙1=phy (ei,aj)calcula es he deg ee o ma ch
be ween he physical and psychological s a us o an indi idual
ei∈Gjand he physical and psychological equi emen s o he
ac i i y aj∈A.
•Func ion 𝜙2=ac (ei,aj)calcula es he ma ch be ween he
pe sonal p e e ences o each indi idual ei∈Gj o he ac i i y
aj∈A.
•Func ion 𝜙3= i (ei,X)calcula es he deg ee o iendship o
an indi idual ei∈Gjwi h o he membe s o he g oup
ek∈Gj∶i≠kconside ing hei iendship ela ionships in
he social ne wo k X.
•Func ion 𝜙4=his (ei,aj,d)is used o penalize he g oup
o med i an indi idual ei∈Gjhas he ac i i y aj∈Ain
his/he his o ical ac i i y p o ile in he las ddays.
Gi en he abo e unc ions, he alue o a g oup is calcula ed as ol-
lows: The pa ame e s 𝛼1,𝛼2,𝛼3, and 𝛼4desc ibed in his equa ion
a e i s ly speci ied wi h he same weigh (0.25):
(Gj)=∑
ei∈Gj
𝛼1⋅phy(ei,aj)+ 𝛼2⋅ac (ei,aj)
+ 𝛼3⋅ i (ei,X)+ 𝛼4⋅his (ei,aj).
(2)
6. EVALUATION
In his sec ion, we p esen some expe imen s o es and alida e
he applica ion. Fi s , we p esen empi ical expe imen s ha we e
ca ied ou in o de o e alua e he pe o mance o he algo i hm
p oposed. Then, we show some esul s abou he opinion o he
ca egi e s ega ding he ool de eloped.
6.1. Expe imen al Se ing
In o de o analyze he pe o mance o he ool ega ding he
di e en c i e ia used o g oup o ma ion, we simula ed se e al
scena ios om he da ase . The alue o each indi idual g oup was
measu ed acco ding o he Equa ion (2). The e o e, he physical and
psychological s a us ma ching o an ac i i y, he pe sonal p e e -
ence o an ac i i y, he iendship ela ionships, and he his o ical
ac i i y p o ile had all o hem he same in luence o compu ing
he alue o a g oup.
Fo each expe imen , we simula ed di e en scena ios in which
indi iduals we e g ouped du ing 30 days. Fo each day, we calcu-
la ed he alue o he g oup s uc u e, conside ing ha each indi id-
ual could ca y ou a single ac i i y pe day. The size o each g oup
anged om h ee o i e people pe g oup, and he numbe o ac i -
i ies was 20. I mus be poin ed ha some ac i i ies could be ca ied
Pd _Folio:669