1510 J. Serrano-Guerrero et al Information Sciences 181 (2011)1503-1516 Resource Recommender System Resource wave 3 Google Wave-based Digital Library Tool Fig 4 Scheme of the google wave-based recommender system for University Digital Libraries. Therefore a resource is defined by a vector VR, apart from other typical data such as a name, a description and an impor tance degree. This last value is firstly fixed by the user who suggests the resource as a member of the wave but can be refixed by the wave administrator s. This value is expressed by a 2-tuple linguistic label bx E S3. 3.3.2. User profiles Both administrators and users have to define their profile before starting to work with the system. the user has to de- scribe his preferences with respect to the same disciplines proposed for the resources in the previous subsection. The way of defining the preferences is the same, the user has to select a 2-tuple linguistic value bx E Sn for each one of the 26 positions of a vector VU. Apart from this vector, the user has to give more information such as his name, family name, email address( Google address to be inserted in the waves), nickname, password, etc. The system also stores a list of waves to which the user is subscribed and a score, a 2-tuple linguistic value bx E S2, for each wave depending on his interest in each wave. This data is provided by the user when he accepts the invitation for participation in the wave. The user also has to define a linguistic threshold value for waves (n) that indicates the minimum value required for informing him about the exis- tence of a wave 3.3.3. Waves The waves present a similar definition with respect to resources and users; a vector W where the administrator has to describe the topics of the wave(see Section 3. 2). But this defin tIon Is ommendations of the system which suggests the most appropriate users who might be interested in and the re- ources that could be interesting according to the theme of the wave.Therefore a resource is defined by a vector VR, apart from other typical data such as a name, a description and an impor￾tance degree. This last value is firstly fixed by the user who suggests the resource as a member of the wave but can be refixed by the wave administrator/s. This value is expressed by a 2-tuple linguistic label bx 2 S3. 3.3.2. User profiles Both administrators and users have to define their profile before starting to work with the system. The user has to de￾scribe his preferences with respect to the same disciplines proposed for the resources in the previous subsection. The way of defining the preferences is the same, the user has to select a 2-tuple linguistic value bx 2 S1 for each one of the 26 positions of a vector VU. Apart from this vector, the user has to give more information such as his name, family name, email address (Google address to be inserted in the waves), nickname, password, etc. The system also stores a list of waves to which the user is subscribed and a score, a 2-tuple linguistic value bx 2 S2, for each wave depending on his interest in each wave. This data is provided by the user when he accepts the invitation for participation in the wave. The user also has to define a linguistic threshold value for waves (c) that indicates the minimum value required for informing him about the exis￾tence of a wave. 3.3.3. Waves The waves present a similar definition with respect to resources and users; a vector VW with 26 positions where the administrator has to describe the topics of the wave (see Section 3.2). But this definition is completed by means of the rec￾ommendations of the system which suggests the most appropriate users who might be interested in the wave and the re￾sources that could be more interesting according to the theme of the wave. Fig. 4. Scheme of the google wave-based recommender system for University Digital Libraries. 1510 J. Serrano-Guerrero et al. / Information Sciences 181 (2011) 1503–1516