J. Serrano-Guerrero et aL Information Sciences 181(2011)1503-1516 1509 B/y№H三理理票,C鲁m ouandExampleengined.com welcome to wavesengin Resources Subject Linguistic label User Preferences Agncultural and Biolog cal sciences Null Agncumural and Biclogica SciencesLow Fig. 3. Interface for inserting a linguistic value for each discipline S=(bo=Null =N, b=Low=L, b2=Medium =M, b 3= High=H, ba=Total=T). S=(co=Null =N, c1=Very-Low =VL, C2=Low=L, C3=More_Less_Low=MLL, Ca=Medium More_Less_high= MLH. C6= High=H, c,= very_High= VH, cs=Total=T ns with respect to the 26 ca gories proposed by ScienceDirect(see Fig 3): (i)Business, Management and Accounting.(i)Chemistry. (ii)Computer Sci- ence, etc. The comparisons between user and user, resource and resource, resource and etc, are based on the use of measures such as the cosine measure. In this case the standard cosine will be the chosen measure based on linguistic values 0(V1,V2)=4g 1.xn)×△-(a2k,xak) where g is the granularity of the used term set, n is the number of terms used to define the vectors (i.e. the number of dis- ciplines)and(oik, apik) is the linguistic value of term k in the user, resource or wave vector Va With this similarity measure we btain a linguistic value in S1 to assess the similarity among two resources, two users, or a resource and a wave, etc. 3.3. Information representation The administrators of the system are people in charge of defining the waves that will exist in the digital library as well ne characteristics of each resource, whereas the user is who inserts his own personal data and main preferences. All nec- essary data to describe each element are discussed below. 3.3.1. Resources Different resources can be found in digital libraries such as books, electronic papers, electronic journals and official dailies [42, 47). An example could be the resources of The Stanford Digital Library Technologies Project. The resources can be classified according to the above-explained classification proposed by ScienceDirect. The adminis- rator who inserts each digital resource has 26 different disciplines to characterize the topics of each resource. The descrip- tion of each resource i is stored in a vector VR with 26 positions, one for each discipline where the administrator can assign a -tuple linguistic label bx E Sn for each position WR={WR1,WR2,…,VRa6} Therefore, each component VRy of the vector VR; indicates the 2-tuple tic importance degree of the dis with j=(1,., 26], with respect to the resource i. Thus the administrator of ive is the person in charge of indic ferent 2-tuple linguistic degrees of each resource that is incorporated i The administrator can insert a new re. source in response to a user who informs about the existence of that useful resource. http://www.info.sciencedirect.c llections 3http://diglib.stanford.edu:8091diglib/pub/resources.shtmlS5 = {b0 = Null = N,b1 = Low = L,b2 = Medium = M,b3 = High = H,b4 = Total = T}. S9 = {c0 = Null = N, c1 = Very_Low = VL, c2 = Low = L, c3 = More_Less_Low = MLL, c4 = Medium = M, c5 = More_Less_High = MLH, c6 = High = H, c7 = Very_High = VH, c8 = Total = T}. The representation of users, waves and resources is a vector [45] that describes their relations with respect to the 26 cat￾egories proposed by ScienceDirect2 (see Fig. 3): (i) Business, Management and Accounting, (ii) Chemistry, (iii) Computer Sci￾ence, etc. The comparisons between user and user, resource and resource, resource and wave, etc., are based on the use of measures such as the cosine measure. In this case the standard cosine will be the chosen measure based on linguistic values: rlðV1; V2Þ ¼ D g  Pn k¼1ðD1 ð#1k; a#1kÞ  D1 ð#2k; a#2kÞÞ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Pn k¼1ðD1 ð#1k; a#1kÞÞ2 q  ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Pn k¼1ðD1 ð#2k; a#2kÞÞ2 q 0 B@ 1 CA; where g is the granularity of the used term set, n is the number of terms used to define the vectors (i.e. the number of dis￾ciplines) and (#ik,a#ik) is the linguistic value of term k in the user, resource or wave vector Vi. With this similarity measure we obtain a linguistic value in S1 to assess the similarity among two resources, two users, or a resource and a wave, etc. 3.3. Information representation The administrators of the system are people in charge of defining the waves that will exist in the digital library as well as the characteristics of each resource, whereas the user is who inserts his own personal data and main preferences. All nec￾essary data to describe each element are discussed below. 3.3.1. Resources Different resources can be found in digital libraries such as books, electronic papers, electronic journals and official dailies [42,47]. An example could be the resources of The Stanford Digital Library Technologies Project.3 The resources can be classified according to the above-explained classification proposed by ScienceDirect. The adminis￾trator who inserts each digital resource has 26 different disciplines to characterize the topics of each resource. The descrip￾tion of each resource i is stored in a vector VRi with 26 positions, one for each discipline where the administrator can assign a 2-tuple linguistic label bx 2 S1 for each position: VRi ¼ fVRi1; VRi2; ... ; VRi26g: Therefore, each component VRij of the vector VRi indicates the 2-tuple linguistic importance degree of the discipline j, with j = {1,..., 26}, with respect to the resource i. Thus the administrator of the wave is the person in charge of indicating the dif￾ferent 2-tuple linguistic degrees of each resource that is incorporated into the wave. The administrator can insert a new re￾source in response to a user who informs about the existence of that useful resource. Fig. 3. Interface for inserting a linguistic value for each discipline. 2 http://www.info.sciencedirect.com/content/backfiles/collections. 3 http://diglib.stanford.edu:8091/diglib/pub/Resources.shtml. J. Serrano-Guerrero et al. / Information Sciences 181 (2011) 1503–1516 1509