Upload
kokome35
View
212
Download
0
Embed Size (px)
Citation preview
7/23/2019 DSpace Definition
1/12
DSpace Definition, Features and Functionality
In March 2000, Hewlett-Packard Company (HP) awarded $1.8 mllon to the MI!"#rare %or an 18-month colla#oraton to #&ld 'pace, a dynamc repotory %orthe ntellect&al o&tp&t n d*tal %ormat o% m<-dcplnary reearch or*an+aton.
HP "a# and MI! "#rare releaed the ytem worldwde on oem#er , 2002,&nder the term o% the /' open o&rce lcene 1, one month a%ter t ntrod&ctona a new erce o% the MI! "#rare. an open o&rce ytem, 'pace now%reely aala#le to other ntt&ton to r&n a-, or to mod%y and e3tend a theyre4&re to meet local need. 5rom the o&tet, HP and MI! de*ned the ytem to #er&n #y ntt&ton other than MI!, and to &pport %ederaton amon* t adopter, n
#oth the techncal and the ocal ene. !he 'pace 5ederaton wll #e e3plored n alater ecton.
o what 'pace6 It an attempt to addre a pro#lem that MI! %ac<y hae #een
e3pren* to the "#rare %or the pat %ew year. %ac<y and other reearcherdeelop reearch materal and cholarly plcaton n ncrean*ly comple3 d*tal%ormat, there a need to collect, preere, nde3 and dtr#&te them7 a tme-con&mn* and e3pene chore %or ndd&al %ac<y and ther department, la#, andcenter to mana*e themele. !he 'pace ytem prode a way to mana*e theereearch materal and plcaton n a pro%eonally mantaned repotory to *ethem *reater #lty and acce#lty oer tme.
'pace wa #< #readth-%rt7 t &pport eery %&ncton that a reearch or*an+atonneed to r&n a prod&cton d*tal repotory erce, #&t a mply a po#le. !he
proect %oc& wa on #&ldn* a prod&cton 4&alty ytem. It complement and wan%l&enced #y preo& reearch n comp&ter cence and d*tal l#rary archtect&re2. 9&r *oal were to #&ld a ytem that7 wo&ld #e mmedately &e%&l at MI!, andhope%&lly at other ntt&ton: co&ld #e e3panded and mproed oer tme: and co&ldere a a plat%orm %or %&t&re reearch. ;th the help o% deeloper at otherntt&ton that adopt 'pace &nder t open o&rce lcene, we wll work to add%eat&re and mproe the d%%erent %&ncton o% the ytem a we learn what &eract&ally want, and how to #et &pport &ch comple3 re4&rement a d*tal
preeraton and d*tal r*ht mana*ement.
'pace de*ned to make partcpaton #y depotor eay. !he ytemnat&ral-&nt o% an ntt&ton that hae dtncte n%ormaton mana*ement need. In thecae o% MI! (a lar*e reearch &nerty) =Comm&nte= are de%ned to #e the chool,department, la#, and center o% the Intt&te. ?ach Comm&nty can adapt the ytemto meet t partc&lar need and mana*e the mon proce tel%.
http://www.dlib.org/dlib/january03/smith/01smith.html#1http://www.dlib.org/dlib/january03/smith/01smith.html#2http://www.dlib.org/dlib/january03/smith/01smith.html#2http://www.dlib.org/dlib/january03/smith/01smith.html#17/23/2019 DSpace Definition
2/12
Figure 1: DSpace information model
Metadata
'pace &e a 4&al%ed 'ln Core metadata tandard %or decr#n* temntellect&ally (pec%cally, the "#rare ;orkn* @ro&p pplcaton Pro%le). 9nlythree %eld are re4&red7 ttle, lan*&a*e, and mon date, all other %eld areoptonal. !here are addtonal %eld %or doc&ment a#tract, keyword, techncalmetadata and r*ht metadata, amon* other. !h metadata dplayed n the temrecord n 'pace, and nde3ed %or #rown* and earchn* the ytem (wthn acollecton, acro collecton, or acro Comm&nte). 5or the 'emnatonIn%ormaton Packa*e ('IP) o% the 9I %ramework, the ytem c&rrently e3portmetadata and d*tal materal n a c&tom AM" chema whle we work wth theM?! B comm&nty to deelop the neceary e3tenon chema %or the techncal
and r*ht metadata a#o&t ar#trary d*tal %ormat.
User Interface
'pace
7/23/2019 DSpace Definition
3/12
!he end-&er or plc nter%ace &pport earch and retreal o% tem #y #rown* orearchn* the metadata (all %eld %or now, and pec%c %eld n the near %&t&re). 9ncean tem located n the ytem, retreal accomplhed #y clckn* a lnk thatca&e the arched materal to #e downloaded to the &er
7/23/2019 DSpace Definition
4/12
Technology platform
'pace wa deeloped to #e open o&rce, and n &ch a way that ntt&ton andor*an+aton wth mnmal reo&rce co&ld r&n t. !he ytem de*ned to r&n onthe DIA plat%orm, and compre other open o&rce mddleware and tool, and
pro*ram wrtten #y the 'pace team. ll or*nal code n the Eaa pro*rammn*lan*&a*e. 9ther pece o% the technolo*y tack ncl&de a relatonal data#aemana*ement ytem (Pot*reF"), a ;e# erer and Eaa erlet en*ne (pache and!omcat, #oth %rom the pache 5o&ndaton), Eena (an G'5 toolkt %rom HP "a#),9ICat %rom 9C"C, and eeral other &e%&l l#rare. ll leera*ed component andl#rare are alo open o&rce o%tware. "#rare are #&ndled where po#le(e3cepton are decr#ed n the ntallaton ntr&cton). !he ytem aala#le ono&rce5or*e , lnked %rom #oth the 'pace n%ormatonal we# te and the HP"a# te .
;hle 'pace open o&rce and %reely aala#le, nether MI! "#rare nor HP o%%er%ormal &pport %or 'pace adopter. It o&r a&mpton that ntt&ton that &e'pace wll hae reo&rce to &e the ytem, ncl&dn* ade4&ate hardware that r&nthe DIA operatn* ytem, and a DIA ytem admntrator to ntall andcon%*&re the ytem J. Mot ntt&ton &n* 'pace wll alo want the erce o%a Eaa pro*rammer who can local+e and c&tom+e %or them, or enhance t, altho&*hth not a#ol&tely neceary to r&n the ytem.
'pace contn&e to #e mproed #y ta%% at HP, the MI! "#rare, and otherntt&ton that adopt t d&rn* the comn* year, MI! wll take repon#lty %or
eal&atn* and rencorporatn* thee mproement nto the man open o&rce ytemaala#le to the plc. Plan %or #&ldn* a more &tana#le open o&rce mantenancetrate*y thro&*h the 'pace 5ederaton wll #e dc&ed later.
System Architecture
http://www.dlib.org/dlib/january03/smith/01smith.html#4http://www.dlib.org/dlib/january03/smith/01smith.html#5http://www.dlib.org/dlib/january03/smith/01smith.html#6http://www.dlib.org/dlib/january03/smith/01smith.html#7http://www.dlib.org/dlib/january03/smith/01smith.html#4http://www.dlib.org/dlib/january03/smith/01smith.html#5http://www.dlib.org/dlib/january03/smith/01smith.html#6http://www.dlib.org/dlib/january03/smith/01smith.html#77/23/2019 DSpace Definition
5/12
Figure 2: DSpace technical architecture
!he 'pace archtect&re a tra*ht%orward three-layer archtect&re, ncl&dn*tora*e, #&ne, and applcaton layer, each wth a doc&mented PI to allow %or%&t&re c&tom+aton and enhancement. !he tora*e layer mplemented &n* the%le ytem, a mana*ed #y Pot*reF" data#ae ta#le. !he #&ne layer wherethe 'pace-pec%c %&nctonalty rede, ncl&dn* the work%low, contentmana*ement, admntraton, and earch and #rowe mod&le. ?ach mod&le ha anPI to allow 'pace adopter to replace or enhance that %&ncton a dered. 5nally,the applcaton layer coer the nter%ace to the ytem7 the we# DI and #atch loader,n partc&lar, #&t alo the 9I &pport and Handle erer %or reoln* pertentdent%er to 'pace tem. !h the layer that wll *et m&ch o% the attenton n
%&t&re releae, a we add we# erce %or new %eat&re (e.*., to &pportnteroperaton wth other ytem) and de%ne 5ederaton erce acro the ran*e o%ntt&ton adoptn* 'pace.
Open Archives Initiative (OAI
!o %&rther t *oal o% &pportn* nteropera#lty wth other 'pace adopter, and wthother d*tal repotore, preprnt, and e-prnt erer, the ytem ha mplemented the
7/23/2019 DSpace Definition
6/12
9pen rche Intate Protocol %or Metadata Haretn* (9I-PMH) 8. 'pace&ed the 9C"C 9ICat K to accomplh th, and c&rrently e3pon* 'ln Coremetadata %or eery tem n the ytem. 5or materal that retrcted to local acce,the tem metadata e3poed to 9I hareter #&t the ytem wll en%orce theretrcton when a &er re4&et the aocated #ttream(). 'pace at MI! ha
recently #een added to the 9I re*try, and a the ytem deployed at otherntt&ton, we ntend to net*ate what added-al&e erce m*ht #e #< on topo% th promn* pece o% n%ratr&ct&re to work acro the 5ederaton. 5or e3ample,we may e3amne the po#lty o% de%nn* and #&ldn* preprnt and e-prntcollecton %or a partc&lar academc dcplne wth ndd&al tem dtr#&tedamon* many ntt&tonally-#aed m<dcplnary repotore, all 9I complant.
!ersistent Identifiers ("andles
9ne *oal o% pertent d*tal repotore that t #e po#le to %nd and retree
depoted tem %ar nto the %&t&re. In partc&lar, t condered cr&cal that ctatonto arched materal, whether %o&nd n prnted artcle or onlne, reman ald %or lon*
perod. !o th end, 'pace choe to mplement CGI handle 10 a the pertentdent%er aocated wth each tem. !he Handle ytemL coer a*nment,mana*ement, and reol&ton o% thee pertent dent%er (or =handle=). ltho&*hCGI ha not re*tered wth the I?!5 %or an o%%cal namepace, handle arecomplant wth the I?!5
7/23/2019 DSpace Definition
7/12
o&r polce may not work well %or other ntt&ton, and wll certanly eole oertme, they may o%%er *&dance to other re*ardn* the depth and #readth o% &e thatho&ld #e condered.
#ollections Scope
t MI!, the or*nal *oal o% 'pace wa to capt&re the %ac<y
7/23/2019 DSpace Definition
8/12
$aculty engagement
!here are eeral way to decr#e the al&e o% an ntt&tonal repotory to the%ac<y who wll contr#&te materal, and the admntraton that wll &pport thee%%ort. nd t crtcal to e3plan thoe #ene%t, and to market the erce, to #oth
contt&ence.
a m<dcplnary repotory that repreent the cholarhp o% MI!, 'pace atMI! howcae the nternatonal promnence o% o&r %ac<y #oth ndd&ally andcollectely. !he nterdcplnary content o% the arche ho&ld attract a wdera&dence than a repotory dedcated to one ndd&al dcplne wo&ld: moreoer t
prode c&rrently lackn* erce to the *rown* #ody o% nterdcplnary reearche%%ort. !he a#lty to dtr#&te reearch re< 4&ckly wll empha+e the c&ttn*-ed*e nat&re o% MI!
7/23/2019 DSpace Definition
9/12
%or an academc comm&nty (e.*., the arA ytem at Cornell Dnerty 1)'pace co&ld #e made to a&tomatcally mt cope o% releant doc&ment to theecentral+ed arche d&rn* the local depot proce.
Transition Team and %usiness plan
5rom the %all o% 2001 &ntl prn* o% 2002, the "#rare %ormed a 'pace !ranton!eam contn* o% proect ta%% and enor l#rary ta%% %rom key department (e.*.,the rche, collecton erce, plc erce, and the ytem department). !h*ro&p wa char*ed wth %*&rn* o&t how to deploy 'pace a a new erce o% theMI! "#rare7 the neceary polce, ta%%n* re4&rement, comm&ncatontrate*e, mana*ement and *oernance tr&ct&re, trann* plan, and operatonalre4&rement. Partcpaton n th *ro&p proed to #e a &e%&l ehcle %or the l#raryta%% to #ecome more %amlar wth the ytem, and dc&on o% thee aro& &ewere nal&a#le to the deelopment o% the prod&cton 'pace erce.
Partcpatn* n the !ranton !eam *ro&p were two enor #&ne con<ant%&nded #y a *rant %rom the ndrew ;. Mellon 5o&ndaton to wrte a %ormal #&ne
plan %or a &tana#le 'pace ytem at MI!. !her work conted o% compln* there< o% the tranton team del#eraton and decon, ncorporatn* the work ntodetaled cot n%ormaton %or ytem operaton, and o&tlnn* po#le reen&e opton.
!he maor concl&on o% th plannn* proce wa that 'pace at MI! wo&ld #eo%%ered a a com#naton o% d+ed core erce (#< nto the "#rare< operatn*
#&d*et), and cot-recoered prem&m erce that wo&ld allow the "#rare to meet
aryn* &n4&e need %or 'pace %rom partc&lar Comm&nte (e.*., e3ceptonalamo&nt o% dk tora*e, atance wth metadata creaton, or coneron o% %le to&pported %ormat). ;th th trate*y we hae n&red that 'pace an a%%orda#le&ndertakn* %or the MI! "#rare wtho&t compromn* the erce that can #eo%%ered 1.
!reservation
Gecent dc&on o% d*tal preeraton %oc& on at leat two leel7 =#tpreeraton=, where a d*tal %le care%&lly preered e3actly a t wa created
wtho&t the l*htet chan*e, and what we
7/23/2019 DSpace Definition
10/12
cae the materal alway kept mmedately &ea#le (ewa#le, playa#le, earcha#le,or whateer yo& co&ld do with itor*nally). 9#o&ly, %&nctonal preeraton themore dera#le leel, #&t t wll come wth a prce.
a comm&nty, o&r &ndertandn* o% %&nctonal d*tal preeraton at an
nteretn* &nct&re7 we know how mportant the need , we know how t can #e doneat an a#tract leel (e.*., %ormat m*raton or comple3 ytem em&laton and o on)./&t %ew ntt&ton hae act&ally had to do %&nctonal preeraton n a prod&ctonettn* on lar*e 4&antte o% hetero*eneo& materal. o we hae ery lttlen%ormaton a#o&t act&al prod&cton trate*e, cot, &er reacton to n%ormaton lo,or how m&ch techncal metadata needed to &pport all o% th.
How doe th all relate to 'pace6 !he ytem capt&re mnmal techncal metadatato &pport d*tal #t preeraton (%le %ormat, M' check&m, creaton date), and
prode decrpte %eld to record more n%ormaton when aala#le. ;th th
metadata and proper prod&cton proced&re (e.*., h*h-4&alty erer and tora*edece, *ood #ack&p and dater recoery plan), 'pace can &pport =#t
preeraton= o that the materal depoted can #e delered to %&t&re &er e3actly at wa or*nally receed. 5or ome d*tal %ormat th may #e the #et optonaala#le>%or e3ample, an e3ec&ta#le pro*ram %or whch no correpondn* o&rcecode wa proded or a %ormat that
7/23/2019 DSpace Definition
11/12
nd&tre dependent on them. I% and when &ch commercal coneron pro*ramemer*e, MI! wll moe thee %ormat nto the =&pported= cate*ory and o%%er%&nctonal preeraton %or them.
The &Space $ederation
nce the ery #e*nnn*, the 'pace proect ntended to make t ytem open o&rceand to actely promote t to other ntt&ton. ;hy6 !here are many reaon %ortakn* th approach7
'eelopn* a crtcal corp& o% content that repreent the ntellect&al o&tp&t o%the world
7/23/2019 DSpace Definition
12/12
Mon* %orward %rom here, there are many, many 4&eton remann*, #&t we %eelthat *reat pro*re ha #een made, and we are ea*er to ee how thn* deelop. tMI! we are ery pleaed and e3cted to hae a plat%orm to #e*n e3plorn* thee&e, #oth wthn the Intt&te and wth other ntt&ton that want to adance thea*enda o% open acce to cholarly n%ormaton and the mana*ement and preeraton
o% d*tal materal. t HP we are e3cted #y the role that 'pace can play a a ehcle%or e3plorn* and deelopn* tandard, and %or on*on* reearch n d*tal aetmana*ement, archal, and preeraton ytem. !o*ether we antcpate that 'pacewll play an mportant role n the %&t&re o% academc l#rare and arche, and welook %orward to prod&cte colla#oraton wth other ntt&ton n th area.
"c#no$ledgements
!he a&thor wo&ld lke to thank o&r ponor7 the Hewlett-PackardMI! llance andthe ndrew ;. Mellon 5o&ndaton. ;e wo&ld alo lke to thank the preo& mem#er
o% the 'pace proect team whoe contr#&ton were nal&a#le, ncl&dn* ?rcCelete, /ll Cattey, 'an Ch&dno, Peter /reton, Peter Carmchael, and Eoyce *.5nally, we wo&ld lke to thank the many collea*&e at HP, MI!, and the "#rare n
partc&lar, who made th proect po#le.