《电子商务 E-business》阅读文献：Motivating and Supporting User Interaction with Recommender Systems

Motivating and Supporting User Interaction with Recommender Systems Andreas w. neumann Institute of Information Systems and Management Universitat Karlsruhe(TH), 76128 Karlsruhe, Germany a neumann@iism. uni-karlsruhe, de http://www.iism.uni-karlsruhe.de/a.neumann Abstract. This contribution reports on the introduction of explicit rec- mender systems at the University Library of Karlsruhe. In March 2006, a rating service and a review service were added to the already ex- isting behavior-based recommender system. Logged-in users can write re- ews and rate all library documents(books, journals, multimedia, etc. reading reviews and inspecting ratings are open to the general public A role system is implemented that supports the submission of different eviews for the same document from one user to different user groups ( students, scientists, etc. ) Mechanism design problems like bias and fre riding are discussed, to address these problems the introduction of in centive systems is described. Usage statistics are given and the question which recommender system supports which user needs best, is covered Summing up, recommender systems are a way to combine the support of library user interaction with information access beyond catalog searches Keywords: Recommender system, rating service, review service, mech- anism design, incentive system. 1 Introduction The general public is lately becoming accustomed with recommender systems of different kinds at various online stores. But scientific libraries, where the profit contribution of a product (library document) is not the first concern and the costumers(library users) are coming due to very different incentives, are defini- tively a not less promising application area. Due to the supply complexity or the evaluation of the quality, scientists and students are more and more inca- pable of efficiently finding relevant literature in conventional database oriented catalog systems and search engines. A common solution to this problem lies in asking peers(see e. g. [10). Recommender systems aggregate knowledge from many peer groups to the level of expert advice services. They bear the poten ial to significantly reduce transaction costs for literature searches by means of their aggregation capabilities. Scientific libraries are in a good strategic posi- tion to become(even more than now) the information centers of the future 7 Turning library online public access catalogs(OPAC) into customer oriented service portals supporting the interaction of the customers is one step to this L Kovacs. N. Fuhr, and C. Meghini(Eds ) ECDL 2007. LNCS 4675. pp. 428-139, 2007 ringer-Verlag Berlin Heidelberg 200

Motivating and Supporting User Interaction with Recommender Systems Andreas W. Neumann Institute of Information Systems and Management, Universit¨at Karlsruhe (TH), 76128 Karlsruhe, Germany a.neumann@iism.uni-karlsruhe.de http://www.iism.uni-karlsruhe.de/a.neumann Abstract. This contribution reports on the introduction of explicit rec￾ommender systems at the University Library of Karlsruhe. In March 2006, a rating service and a review service were added to the already ex￾isting behavior-based recommender system. Logged-in users can write re￾views and rate all library documents (books, journals, multimedia, etc.); reading reviews and inspecting ratings are open to the general public. A role system is implemented that supports the submission of different reviews for the same document from one user to different user groups (students, scientists, etc.). Mechanism design problems like bias and free riding are discussed, to address these problems the introduction of in￾centive systems is described. Usage statistics are given and the question, which recommender system supports which user needs best, is covered. Summing up, recommender systems are a way to combine the support of library user interaction with information access beyond catalog searches. Keywords: Recommender system, rating service, review service, mech￾anism design, incentive system. 1 Introduction The general public is lately becoming accustomed with recommender systems of different kinds at various online stores. But scientific libraries, where the profit contribution of a product (library document) is not the first concern and the costumers (library users) are coming due to very different incentives, are defini￾tively a not less promising application area. Due to the supply complexity or the evaluation of the quality, scientists and students are more and more inca￾pable of efficiently finding relevant literature in conventional database oriented catalog systems and search engines. A common solution to this problem lies in asking peers (see e. g. [10]). Recommender systems aggregate knowledge from many peer groups to the level of expert advice services. They bear the poten￾tial to significantly reduce transaction costs for literature searches by means of their aggregation capabilities. Scientific libraries are in a good strategic posi￾tion to become (even more than now) the information centers of the future [7]. Turning library online public access catalogs (OPAC) into customer oriented service portals supporting the interaction of the customers is one step to this L. Kov´acs, N. Fuhr, and C. Meghini (Eds.): ECDL 2007, LNCS 4675, pp. 428–439, 2007. c Springer-Verlag Berlin Heidelberg 2007

Motivating and Supporting User Interaction with Recommender Systems 429 goal. Valid and credible information is a scarce resource[ 20. Information con- sumes the attention of its recipients. Hence a wealth of information creates a poverty of attention and a need to allocate that attention efficiently among the overabundance of information sources that might consume it. "211 The more general term"recommender system"was coined by Resnick and Varian to better describe the action than the more narrow " collaborative fil- tering"16. A recommender system reads observed user behavior or opinions from users as input, then aggregates and directs the resulting recommendations to appropriate recipients. Recommender systems can be classified into two dif- ferent main categories. An implicit recommender system is based on behavioral usage data like purchases, digital library catalog inspections, or lending data An explicit recommender system directly asks the users for their opinions on certain objects. A more technical classification with a focus on applications in e-commerce can be found in [18 and [ 19. For a more up-to-date overview on recommender systems e. g. see Adomavicius and Tuzhilin [1]. In Geyer-Schulz et al. 5 an early application of recommender systems including group-specific services in e-learning is presented. Herlocker et al. 9 deals with the technical evaluation of recommender systems The focus of this paper lies on the experiences with motivation and support of interaction between library users at the University Library of Karlsruhe. First the introduced recommender systems are described, then mechanism design is discussed to address motivational problems. Finally, general lessons learned from integrating different recommender systems into large existing legacy library ap- plications are summarized and the evaluation of such systems is discussed All in this paper presented recommender systems are fully operational services accessible by the general public. For further information on how to use these see "parTicipate!"athttp://reckvk.em.uni-karlsruhe.de/.Inanswertostrong privacy concerns among students and scientists all portrayed recommender ser- vices are object-centered. They do not classify the users by observation or asking them for their interest, but they classify and gather data on the documents of a library. Figure 1 shows a cutout of the detailed document inspection page of [13] in the OPac of the University Library of Karlsruhe. The behavior-based ser vice is accessibly by clicking on "Empfehlungen"(Recommendations), the rating service by "Bewertung abgeben"(Submit rating) or direct inspection of"Bew ertung des Titels nach Nutzergruppen"(Ratings of the titles by user group) and finally the review service by "Rezension schreiben"(Write review),"Rezen- sionen anzeigen"(Inspect reviews), and"Meine Rezensionen"(My reviews). All systems are programmed in Perl or PHP(or a combination of both), use Post- greSQL databases, and are running on Linux servers 2 Behavior-Based Recommender service Behavior-based recommender services are observing the behavior of users and thereby implicitly collecting information about the objects the users are inspect g. The necessary homogeneity of a group of users in this case is granted by

Motivating and Supporting User Interaction with Recommender Systems 429 goal. Valid and credible information is a scarce resource [20]. Information con￾sumes the attention of its recipients. “Hence a wealth of information creates a poverty of attention and a need to allocate that attention efficiently among the overabundance of information sources that might consume it.” [21] The more general term “recommender system” was coined by Resnick and Varian to better describe the action than the more narrow “collaborative fil￾tering” [16]. A recommender system reads observed user behavior or opinions from users as input, then aggregates and directs the resulting recommendations to appropriate recipients. Recommender systems can be classified into two dif￾ferent main categories. An implicit recommender system is based on behavioral usage data like purchases, digital library catalog inspections, or lending data. An explicit recommender system directly asks the users for their opinions on certain objects. A more technical classification with a focus on applications in e-commerce can be found in [18] and [19]. For a more up-to-date overview on recommender systems e. g. see Adomavicius and Tuzhilin [1]. In Geyer-Schulz et al. [5] an early application of recommender systems including group-specific services in e-learning is presented. Herlocker et al. [9] deals with the technical evaluation of recommender systems. The focus of this paper lies on the experiences with motivation and support of interaction between library users at the University Library of Karlsruhe. First, the introduced recommender systems are described, then mechanism design is discussed to address motivational problems. Finally, general lessons learned from integrating different recommender systems into large existing legacy library ap￾plications are summarized and the evaluation of such systems is discussed. All in this paper presented recommender systems are fully operational services accessible by the general public. For further information on how to use these see “Participate!” at http://reckvk.em.uni-karlsruhe.de/. In answer to strong privacy concerns among students and scientists all portrayed recommender ser￾vices are object-centered. They do not classify the users by observation or asking them for their interest, but they classify and gather data on the documents of a library. Figure 1 shows a cutout of the detailed document inspection page of [13] in the OPAC of the University Library of Karlsruhe. The behavior-based ser￾vice is accessibly by clicking on “Empfehlungen” (Recommendations), the rating service by “Bewertung abgeben” (Submit rating) or direct inspection of “Bew￾ertung des Titels nach Nutzergruppen” (Ratings of the titles by user group), and finally the review service by “Rezension schreiben” (Write review), “Rezen￾sionen anzeigen” (Inspect reviews), and “Meine Rezensionen” (My reviews). All systems are programmed in Perl or PHP (or a combination of both), use Post￾greSQL databases, and are running on Linux servers. 2 Behavior-Based Recommender Service Behavior-based recommender services are observing the behavior of users and thereby implicitly collecting information about the objects the users are inspect￾ing. The necessary homogeneity of a group of users in this case is granted by

A.W. Neumann Rezension schreiben Bewertung abgeben Rezensionen anzeige Meine Rezensionen Empfehlungen Bewertung des Titels nach 食★★★★ Studenten:35(4Bew) ★☆★★ o Mtarbeiter:445Bew Fig 1. Recommender start interface on a document's detailed inspection page sion schreiben- Write review: Bewertung abgeben- Submit rating: Rezensionen nzeigen-Inspect reviews: Meine Rezensionen- My reviews: Empfehlungen mendations: Bewertung des Titels nach Nutzergruppen- Ratings of the titles by user group the principle of self-selection [17, 22 In a library setting usage behavior can be observed at different stages: detailed inspections of documents in the OPAc ordering paper documents from the magazine, ordering paper documents that are currently lent, and finally picking-up a paper document or downloading a file from the digital library. The main concern for the data selection is bias. It can be shown that lending and ordering data is highly biased, since e.g. stu- dents very often do not order the book they are mostly interested in, because most likely it is already lent, but actually their consideration-set only includes documents that they will be able to obtain timely before the corresponding ex amination. In marketing several conceptual models which describe a sequence of sets(e.g. total set 2 awareness set 2 consideration set 2 choice set(11 p. 153)) have been developed to describe such situations [14, 23. For this reason the behavior-based recommender service at the University Library of Karlsruhe based on anonymized OPAC searches(hits on document inspection pages)and ot on lending data. Due to transaction costs the detailed inspection of docu- ments in the OPAC of a library can be put on a par with a purchase incidence in a consumer store setting. A market basket consists of all documents that have been co-inspected by one anonymous user within one session. To answer the question, which co-inspections occur non-random, an algorithm based on calculating inspection frequency distribution functions following a logarithmic series distribution(LSD)is applied 6. Such a recommender system is opera- tional at the OPac of the University Library of Karlsruhe in a first version since

430 A.W. Neumann Fig. 1. Recommender start interface on a document’s detailed inspection page. Rezen￾sion schreiben – Write review; Bewertung abgeben – Submit rating; Rezensionen anzeigen – Inspect reviews; Meine Rezensionen – My reviews; Empfehlungen – Recom￾mendations; Bewertung des Titels nach Nutzergruppen – Ratings of the titles by user group. the principle of self-selection [17,22]. In a library setting usage behavior can be observed at different stages: detailed inspections of documents in the OPAC, ordering paper documents from the magazine, ordering paper documents that are currently lent, and finally picking-up a paper document or downloading a file from the digital library. The main concern for the data selection is bias. It can be shown that lending and ordering data is highly biased, since e.g. stu￾dents very often do not order the book they are mostly interested in, because most likely it is already lent, but actually their consideration-set only includes documents that they will be able to obtain timely before the corresponding ex￾amination. In marketing several conceptual models which describe a sequence of sets (e. g. total set ⊇ awareness set ⊇ consideration set ⊇ choice set ([11], p. 153)) have been developed to describe such situations [14,23]. For this reason the behavior-based recommender service at the University Library of Karlsruhe is based on anonymized OPAC searches (hits on document inspection pages) and not on lending data. Due to transaction costs the detailed inspection of docu￾ments in the OPAC of a library can be put on a par with a purchase incidence in a consumer store setting. A market basket consists of all documents that have been co-inspected by one anonymous user within one session. To answer the question, which co-inspections occur non-random, an algorithm based on calculating inspection frequency distribution functions following a logarithmic series distribution (LSD) is applied [6]. Such a recommender system is opera￾tional at the OPAC of the University Library of Karlsruhe in a first version since

Motivating and Supporting User Interaction with Recommender Systems 431 nmen mit folgenden Titeln auger 2006 1. Hgh Performance Linux Clusters / Sloan, Joseph D, 2005. (16) 围园 verteilte Programmierung /Rauber, Thomas: RUnger, Gudula, 2000. (10)a E d parallel computing 4. Using MPl/ Gropp, D: Lusk, Ewing L; Skiellum, Anthony, 1999, (9) 7) 回回回回回回 9. Beowulf cluster computing with Linux/ Sterling, Thamas Lawrence, 2002.(6 11. Custern mit Hintergrundwssen /Hotho, Andreas, 2004,(6) 2. C und Linux/ Grafe, Martin, 2005. (3) Fig. 2. Recommendation list of"Cluster computing" by Bauke and Mertens. The num- ber of co-inspections is given in brackets after each title June 2002 8 and in the current web service version(facilitating WSDL, XML and SOAP) since January 2006 Figure 2 shows the recommendation list of"Cluster computing"by Bauke nd Mertens(cut-out from the web page). The number of co-inspections is given in brackets after each title. Documents with less than three co-inspections have been rated by the lsd test to be not significantly related to this book. Since the usage distribution of documents in nearly every library is highly skewed(newer documents, or documents to topics that interest a large part of the overall library users, in general are more requested), many recommendations will be generated for documents that are used often while seldom used documents have fewer or no recommendations. Recommendations are updated daily. Of the 929 637 doc uments in the catalog, 192 647 documents have lists with recommendations, a total of 2 843017 recommendations exist. Because of the skewness, the coverage of actual detailed document inspections is 74.9%(much higher than the cover age of the complete catalog). So the probability that recommendations exist for a document a user is currently interested in is 0.749(status of 2007-03-19).A user survey asking the library users"I consider the recommendation service in general "on a Likert scale from 1(very bad) to 5(very good) yielded a mean of 4. 1 from 484 votes between 2005-03-21 and 2006-03-06. This type of recom- mender service is best suited to users trying to find standard literature or further standard readings of a field corresponding to the document they are currently nspecting. Although it does not support the direct interaction(communication) between customers, everybody using the service profits from the actions of other library users An e-mail notification service was added at a later stage. Users with a library account receive an e-mail including a direct link to the recommendation page if

Motivating and Supporting User Interaction with Recommender Systems 431 Fig. 2. Recommendation list of “Cluster computing” by Bauke and Mertens. The num￾ber of co-inspections is given in brackets after each title. June 2002 [8] and in the current web service version (facilitating WSDL, XML and SOAP) since January 2006. Figure 2 shows the recommendation list of “Cluster computing” by Bauke and Mertens (cut-out from the web page). The number of co-inspections is given in brackets after each title. Documents with less than three co-inspections have been rated by the LSD test to be not significantly related to this book. Since the usage distribution of documents in nearly every library is highly skewed (newer documents, or documents to topics that interest a large part of the overall library users, in general are more requested), many recommendations will be generated for documents that are used often while seldom used documents have fewer or no recommendations. Recommendations are updated daily. Of the 929637 doc￾uments in the catalog, 192647 documents have lists with recommendations, a total of 2 843017 recommendations exist. Because of the skewness, the coverage of actual detailed document inspections is 74.9% (much higher than the cover￾age of the complete catalog). So the probability that recommendations exist for a document a user is currently interested in is 0.749 (status of 2007-03-19). A user survey asking the library users “I consider the recommendation service in general” on a Likert scale from 1 (very bad) to 5 (very good) yielded a mean of 4.1 from 484 votes between 2005-03-21 and 2006-03-06. This type of recom￾mender service is best suited to users trying to find standard literature or further standard readings of a field corresponding to the document they are currently inspecting. Although it does not support the direct interaction (communication) between customers, everybody using the service profits from the actions of other library users. An e-mail notification service was added at a later stage. Users with a library account receive an e-mail including a direct link to the recommendation page if

432 A.W. Neumann new recommendations appear for a previously specified document. The usage of this service didn't meet the first expectations. Users seem to be skeptic about any service that tries to grab their attention(like spam mails)at times when they didn't even visit the library. To overcome this problem it is planned to extend this notification service in the near future to support RSS feed techniques. Thereby, each user can decide within the RSs reader when to poll the service. Further on, this way it is no longer connected to existing user accounts, but opened personalized service to the general public 3 Explicit Recommender Systems Two different kinds of explicit recommender systems are online at the University Library of Karlsruhe since March 2006, a rating service and a review service. To prevent fraudulent use, submitting ratings and reviews is possible only for logged- in users. These services differ from most other systems(e. g. Amazon coms) by means of user and target groups and strict separation of ratings and reviews Currently three different user groups exist: students(Studenten), university staff Mitarbeiter), and others(Externe)not directly associated with the universit While one could easily come up with more elaborate user classifications, these vices. They are checked(and afterwards tracked over time)by th ag these ser- three groups have been used by the library for many years prece e library for each user before handing out the library card. The guarantee of correctness made this user classification the(pragmatic) choice of approach for a library with an existing base of approximately 24 400 active users 3.1 Rating Service This service allows logged-in users so submit a numerical rating for a document on a Likert scale from 1 (very bad) to 5(very good). Every user can submit only one rating per document. The ratings are aggregated for each user group separately and are shown in numerical form(average rating, number of ratings) as well as an enlightened-star-graphic on the detailed document inspection page In figure I we see 4 ratings from students(Studenten) with an average of 3.5 and 5 ratings from university staff(Mitarbeiter) with an average 4.4. Thus, at a first glance[13 seems to be an overall good book, even more praised by scientists than by students Figure 3 shows the overall number of ratings online from 2006-03-03 to 2007 03-19. One large draw back of the current setup is known. Users searching the catalog are normally not logged-in until they want to order a paper copy of a document not freely available right now. To submit ratings they have to first log-in. This hurdle seems to have a huge influence on the number or submitted ratings, although it should have a very positive influence on the quality of the ratings. This service is best suited to get a first quick estimation of the overall quality of a document within certain user groups

432 A.W. Neumann new recommendations appear for a previously specified document. The usage of this service didn’t meet the first expectations. Users seem to be skeptic about any service that tries to grab their attention (like spam mails) at times when they didn’t even visit the library. To overcome this problem it is planned to extend this notification service in the near future to support RSS feed techniques. Thereby, each user can decide within the RSS reader when to poll the service. Further on, this way it is no longer connected to existing user accounts, but opened as a personalized service to the general public 3 Explicit Recommender Systems Two different kinds of explicit recommender systems are online at the University Library of Karlsruhe since March 2006, a rating service and a review service. To prevent fraudulent use, submitting ratings and reviews is possible only for logged￾in users. These services differ from most other systems (e. g. Amazon.com’s) by means of user and target groups and strict separation of ratings and reviews. Currently three different user groups exist: students (Studenten), university staff (Mitarbeiter), and others (Externe) not directly associated with the university. While one could easily come up with more elaborate user classifications, these three groups have been used by the library for many years preceding these ser￾vices. They are checked (and afterwards tracked over time) by the library for each user before handing out the library card. The guarantee of correctness made this user classification the (pragmatic) choice of approach for a library with an existing base of approximately 24400 active users. 3.1 Rating Service This service allows logged-in users so submit a numerical rating for a document on a Likert scale from 1 (very bad) to 5 (very good). Every user can submit only one rating per document. The ratings are aggregated for each user group separately and are shown in numerical form (average rating, number of ratings) as well as an enlightened-star-graphic on the detailed document inspection page. In figure 1 we see 4 ratings from students (Studenten) with an average of 3.5 and 5 ratings from university staff (Mitarbeiter) with an average 4.4. Thus, at a first glance [13] seems to be an overall good book, even more praised by scientists than by students. Figure 3 shows the overall number of ratings online from 2006-03-03 to 2007- 03-19. One large drawback of the current setup is known. Users searching the catalog are normally not logged-in until they want to order a paper copy of a document not freely available right now. To submit ratings they have to first log-in. This hurdle seems to have a huge influence on the number or submitted ratings, although it should have a very positive influence on the quality of the ratings. This service is best suited to get a first quick estimation of the overall quality of a document within certain user groups

Motivating and Supporting User Interaction with Recommender Systems 433 Number of Ratings and Reviews ratings 8 revIews 2006-03-012006-05-012006-07-012006-09-012006-11-012007-01-012007-03-01 Fig 3. Number of ratings and reviews online for the general public in the OPAc from 2006-03-03to2007-03-19 www.ubka.uni-karisruhe.de.MeineRezens cht nach auden sichuan Nlar Recension wy books on this wpe annoy the reader because of their Mahat (rm, Schritt fur schmitt db kit 88 C nicht zugeardnet c F Mitarbeiter C Externe NPE(Number of Review) 25n57nscht sichtbar) Service management and marketing/GrunrDos, Christian Rezensent Neumann, Andreas(Mitarbeiter) Freigabe欲us Fig. 4. Editing page for reviews. The author has to choose a target group(Zielgruppe) and his own degree of anonymity(Anonymitatsgrad)

Motivating and Supporting User Interaction with Recommender Systems 433 0 50 100 150 200 Number of Ratings and Reviews Number 2006−03−01 2006−05−01 2006−07−01 2006−09−01 2006−11−01 2007−01−01 2007−03−01 ratings reviews Fig. 3. Number of ratings and reviews online for the general public in the OPAC from 2006-03-03 to 2007-03-19 Fig. 4. Editing page for reviews. The author has to choose a target group (Zielgruppe) and his own degree of anonymity (Anonymit¨atsgrad)

A.W. Neumann 3. 2 Review service The review service manages document reviews written by library users. Figure 4 shows the editing page for reviews. Every logged-in user is allowed to submit four different reviews for each of the library's documents, one addressed to each of the three user groups(Zielgruppe the target groups of the review) and a fourth one not assigned to any user group. This offers the possibility to focus the reviews on the specific needs of the target groups. Parts of a book may be suited very well for a specific course(target group of students), while other aspect e. g. the cited literature are mostly valued by scientists. Reviews can be written and saved within the system over several sessions, only after explicitly releasing a review, it shows up in the OPAC. The user can choose for each review, if it published anonymously or under his real name. Writers are informed about guidelines for reviews, but no further checks from library staff is included in the workfow of submitting a review. Offending reviews can be reported to the library by every user. Since the writer of every review is known at least to the library, such reviews could be deleted and the writers contacted. No such case has been reported so far, although some users have deleted some of their own reviews(confer figure 3) ka. uni.karlsruhe. de. Rezensionen zum Titel. Mozilla Firefox status des Autors ch Sidemen ch Arbeiter dbkit 8 ture, and Corporate ConD 时m29地 Fig. 5. Browsing page for reviews. The author (Autor) chose to stay anonymous (XXXX)but belongs to the staff group(Mitarbeiter), the target group(Zielgruppe) students(Studenten), it has been rated(ezensionsbewertung) one time by students and two times by university staff(see stars). On the left hand side various sorting criteria(up and down) for reviews exist: reviewer(Rezensent), date(Datum), ratings (Bewertung) from the different user groups, reviewer group (Status des Rezensenten) as well as target group

434 A.W. Neumann 3.2 Review Service The review service manages document reviews written by library users. Figure 4 shows the editing page for reviews. Every logged-in user is allowed to submit four different reviews for each of the library’s documents, one addressed to each of the three user groups (Zielgruppe – the target groups of the review) and a fourth one not assigned to any user group. This offers the possibility to focus the reviews on the specific needs of the target groups. Parts of a book may be suited very well for a specific course (target group of students), while other aspects e. g. the cited literature are mostly valued by scientists. Reviews can be written and saved within the system over several sessions, only after explicitly releasing a review, it shows up in the OPAC. The user can choose for each review, if it is published anonymously or under his real name. Writers are informed about guidelines for reviews, but no further checks from library staff is included in the workflow of submitting a review. Offending reviews can be reported to the library by every user. Since the writer of every review is known at least to the library, such reviews could be deleted and the writers contacted. No such case has been reported so far, although some users have deleted some of their own reviews (confer figure 3). Fig. 5. Browsing page for reviews. The author (Autor) chose to stay anonymous (XXXX) but belongs to the staff group (Mitarbeiter), the target group (Zielgruppe) is students (Studenten), it has been rated (Rezensionsbewertung) one time by students and two times by university staff (see stars). On the left hand side various sorting criteria (up and down) for reviews exist: reviewer (Rezensent), date (Datum), ratings (Bewertung) from the different user groups, reviewer group (Status des Rezensenten), as well as target group

Motivating and Supporting User Interaction with Recommender Systems 435 Figure 5 shows the browsing page for reviews. A rating service analog to the one described in section 3. 1 is available on the level of reviews. By means of this a first impression of the quality of certain reviews can be assessed without reading them. Reviews can be browsed and sorted by different criteria: reviewer, date, average ratings of the three user groups respectively, user group of the reviewer, and target group. By means of this service more detailed information about the content, the quality and the adequacy of a document for certain tasks (like preparation for an examination) can be assessed, even if the full text of the document is not available online Inspection of this information on the other hand takes significantly longer than with the previously described systems. When searching the full text of all reviews for keywords, it can be used as a user generated indexing of the library catalog. At 2007-03-19 26 reviews(see figure 3) and 11 ratings of reviews are online. The reasons behind these numbers are discussed in the following sections 4 Mechanism Design Problems,, and solutions Motivating users to write reviews or rate documents in a digital library is a game of(static) mechanism design, a special class of games of incomplete in formation. See e.g. Game Theory"by Fudenberg and Tirole [4 pp for an introduction. By determining the structure of the digital library and e corresponding recommender services the operator of the library chooses the mechanism that maximizes his desired outcome. Here, the players are all library sers and the desired outcome is a large number of high quality(implicit and explicit)recommendations. The following mechanism design problems are most dominant in the described applications Free-riding. Observing recommendations is highly valued, but due to transac tion costs few users actually are willing to produce them Bias. Conscious or unconscious prejudice. E. g. a book author favors his product to the ones of competitors Credibility. Are recommendations mixed with sales promotion or advertise- ments? Privacy vs. recognition of good cooperation. To laud users with exemp- ry cooperation you need their allowance to recognize them Positive/Negative feedback effects. The first good or bad recommendation may lead to further good or bad recommendations respectively(path Economies of scale. The more contributing users(and thus recommendations) a system has, the more useful it is and thereby attracting even more users. o solve these problems a suited incentive systems has to be implemented Recommendations are no standard consumer goods thus needing a special user motivation approach 2. Motivation can be intrinsic or extrinsic. Extrinsic mo- tivation is generated e. g by payments or public commendation. Compensations not only fulfill the purpose of inducing effort on the existing user group but

Motivating and Supporting User Interaction with Recommender Systems 435 Figure 5 shows the browsing page for reviews. A rating service analog to the one described in section 3.1 is available on the level of reviews. By means of this a first impression of the quality of certain reviews can be assessed without reading them. Reviews can be browsed and sorted by different criteria: reviewer, date, average ratings of the three user groups respectively, user group of the reviewer, and target group. By means of this service more detailed information about the content, the quality and the adequacy of a document for certain tasks (like preparation for an examination) can be assessed, even if the full text of the document is not available online. Inspection of this information on the other hand takes significantly longer than with the previously described systems. When searching the full text of all reviews for keywords, it can be used as a user generated indexing of the library catalog. At 2007-03-19 26 reviews (see figure 3) and 11 ratings of reviews are online. The reasons behind these numbers are discussed in the following sections. 4 Mechanism Design Problems . . . and Solutions Motivating users to write reviews or rate documents in a digital library is a game of (static) mechanism design, a special class of games of incomplete in￾formation. See e. g. “Game Theory” by Fudenberg and Tirole [4] pp. 243–318 for an introduction. By determining the structure of the digital library and the corresponding recommender services the operator of the library chooses the mechanism that maximizes his desired outcome. Here, the players are all library users and the desired outcome is a large number of high quality (implicit and explicit) recommendations. The following mechanism design problems are most dominant in the described applications: Free-riding. Observing recommendations is highly valued, but due to transac￾tion costs few users actually are willing to produce them. Bias. Conscious or unconscious prejudice. E.g. a book author favors his product to the ones of competitors. Credibility. Are recommendations mixed with sales promotion or advertise￾ments? Privacy vs. recognition of good cooperation. To laud users with exemp￾lary cooperation you need their allowance to recognize them. Positive/Negative feedback effects. The first good or bad recommendation may lead to further good or bad recommendations respectively (path dependency). Economies of scale. The more contributing users (and thus recommendations) a system has, the more useful it is and thereby attracting even more users. To solve these problems a suited incentive systems has to be implemented. Recommendations are no standard consumer goods thus needing a special user motivation approach [2]. Motivation can be intrinsic or extrinsic. Extrinsic mo￾tivation is generated e.g. by payments or public commendation. Compensations not only fulfill the purpose of inducing effort on the existing user group but

A.W. Neumann also aiding the selection of appropriate new users 15. On the other hand, when offering compensations, intrinsic motivation is often displaced by extrinsic moti- vation. So, once you offer compensations e. g. in form of free book donations to the best reviewers, you scare away some users, that were willing to contribut out of altruism or their implicit membership to the scientific community before Unfortunately, it has been shown that experiments to measure these motivations correctly are very hard to accomplish 12 In e-commerce applications shilling of recommender systems is often a mo tivation. The possibility to submit anonymously (or with fake accounts only requiring an e-mail address) ratings and reviews for one's own products to boost sales leads to significantly more submissions. This mechanism is less dominant in a library setting. The more restrictive the submission process is handled, the less submissions can be expected. The recommender systems at the University Library of Karlsruhe in the current first implementation are very restrictive in the area of the accepted user group and the anonymity towards the system ad ministrator Lessening the restrictions may lead to more submissions with the drawback of a higher rate of biased ratings and reviews 3. In general, mechanism design problems are of less concern with behavior systems. ng consciously in a well implemented system(including web robot prevention) has very high transaction costs and therefore is mostly unattractive in a library setting, and finally all users of the OPAC (regardless of their interest in rec- ommendations) contribute to the recommender system and thereby helping to scale it up. Achieving the critical mass is the most important goal for stand- alone recommender systems(cold start problem) but is less indispensable to life for systems that are placed as value-added services to already high frequented information centers like digital library OPACs. The credibility in the academic environment comes to a large part from the institution to which the library be- longs. If promotions or advertisements of any kind within the OPAc exist, a user should perceive a clear separation between these and the recommender system This is often not the case in e-commerce applications like e. g. Amazon. com, where products with a high contribution to profit are placed by product man- gers next to real recommendations from other customers. Recognition of good cooperation within explicit recommender systems can be measured by reputa tion systems(for credence goods e. g. see 3). A user point account tracks useful behavior(credit)and undesirable behavior(deduction of points). To keep users motivated an automatic discounting(decrease of points) over time is necessary The quality of a review e.g. can be measured by the ratings of other users for this review 5 Conclusions and Further research Scientific libraries hold a good strategic position to become digital information centers. Such information centers need to support library user interaction as well

436 A.W. Neumann also aiding the selection of appropriate new users [15]. On the other hand, when offering compensations, intrinsic motivation is often displaced by extrinsic moti￾vation. So, once you offer compensations e. g. in form of free book donations to the best reviewers, you scare away some users, that were willing to contribute out of altruism or their implicit membership to the scientific community before. Unfortunately, it has been shown that experiments to measure these motivations correctly are very hard to accomplish [12]. In e-commerce applications shilling of recommender systems is often a mo￾tivation. The possibility to submit anonymously (or with fake accounts only requiring an e-mail address) ratings and reviews for one’s own products to boost sales leads to significantly more submissions. This mechanism is less dominant in a library setting. The more restrictive the submission process is handled, the less submissions can be expected. The recommender systems at the University Library of Karlsruhe in the current first implementation are very restrictive in the area of the accepted user group and the anonymity towards the system ad￾ministrator. Lessening the restrictions may lead to more submissions with the drawback of a higher rate of biased ratings and reviews. In general, mechanism design problems are of less concern with behavior￾based recommender systems. Free-riding is almost not possible, to create bias consciously in a well implemented system (including web robot prevention) has very high transaction costs and therefore is mostly unattractive in a library setting, and finally all users of the OPAC (regardless of their interest in rec￾ommendations) contribute to the recommender system and thereby helping to scale it up. Achieving the critical mass is the most important goal for stand￾alone recommender systems (cold start problem) but is less indispensable to life for systems that are placed as value-added services to already high frequented information centers like digital library OPACs. The credibility in the academic environment comes to a large part from the institution to which the library be￾longs. If promotions or advertisements of any kind within the OPAC exist, a user should perceive a clear separation between these and the recommender system. This is often not the case in e-commerce applications like e. g. Amazon.com, where products with a high contribution to profit are placed by product man￾agers next to real recommendations from other customers. Recognition of good cooperation within explicit recommender systems can be measured by reputa￾tion systems (for credence goods e. g. see [3]). A user point account tracks useful behavior (credit) and undesirable behavior (deduction of points). To keep users motivated an automatic discounting (decrease of points) over time is necessary. The quality of a review e. g. can be measured by the ratings of other users for this review. 5 Conclusions and Further Research Scientific libraries hold a good strategic position to become digital information centers. Such information centers need to support library user interaction as well

Motivating and Supporting User Interaction with Recommender Systems 437 as information access beyond catalog searches. Recommender systems are a way to combine both. Different recommender systems support different user needs (e. g. finding standard literature or finding a specialized document for a specific topic). To amplify the described services the derived information is going to be stronger connected in the future. On one hand, e. g. the rating data can be used to further filter the behavior-based recommendations, on the other hand a different graph-based visualization approach that portrays the heterogeneous data from the different systems within one view is developed. Another way is a market-based approach to decide which information from which system should be offered to the user. The principle design of such a marketplace is described All presented recommender systems are becoming regular OPAC features at e University Library of Karlsruhe. The introduction of the implicit recom- mender services is conducted in several steps. The first step comprised the te nical development and launch of the services in the form described by this paper To measure the intrinsic motivation of the users and to find the main obstacles for the users within the system, no technical incentive system like user point accounts was included, neither were any users directly asked to write reviews or give ratings. Although a lot of positive feedback for the systems itself was received, the free-riding problem can be hold responsible for the overall low in formation users have been put into the system. To overcome this situation, in the next steps the following is planned. First, students will be asked to write reviews on literature they are using for seminars to increase the number of quality r views. Second, a reputation systems(list of best reviewers, best reviews, etc )will be included and will be accompanied at an even later stage by a compensation system to raise extrinsic motivation Throughout all steps the evaluation of the quality of the ratings and reviews are of concern as well. Currently, the quality of reviews is measured by the rat ings of reviews. No objective metric exists to measure the quality of scientifie documents in an absolute way, the metric always depends on the function document has to fulfill for a specific user. Once a reasonable number of sub- missions of ratings of documents exists, these ratings could be compared with data from other systems like Amazon. com, data from citation indices might correlate with ratings from scientists, and an evaluation by experts (lectur- ers, librarians, etc. )could lead to further insights as well. The most reasonable way to measure the effectiveness of the systems lies in observing the usage of the systems and asking the users, if the recommender systems(and thereby other users) helped them to find the right literature for the task they had in mn Acknowledgments. The author gratefully acknowledge the funding of the project "Recommender Systems for Meta Library Catalogs" by the Deutsche Forschungsgemeinschaft

Motivating and Supporting User Interaction with Recommender Systems 437 as information access beyond catalog searches. Recommender systems are a way to combine both. Different recommender systems support different user needs (e. g. finding standard literature or finding a specialized document for a specific topic). To amplify the described services the derived information is going to be stronger connected in the future. On one hand, e. g. the rating data can be used to further filter the behavior-based recommendations, on the other hand a different graph-based visualization approach that portrays the heterogeneous data from the different systems within one view is developed. Another way is a market-based approach to decide which information from which system should be offered to the user. The principle design of such a marketplace is described in [24]. All presented recommender systems are becoming regular OPAC features at the University Library of Karlsruhe. The introduction of the implicit recom￾mender services is conducted in several steps. The first step comprised the tech￾nical development and launch of the services in the form described by this paper. To measure the intrinsic motivation of the users and to find the main obstacles for the users within the system, no technical incentive system like user point accounts was included, neither were any users directly asked to write reviews or give ratings. Although a lot of positive feedback for the systems itself was received, the free-riding problem can be hold responsible for the overall low in￾formation users have been put into the system. To overcome this situation, in the next steps the following is planned. First, students will be asked to write reviews on literature they are using for seminars to increase the number of quality re￾views. Second, a reputation systems (list of best reviewers, best reviews, etc.) will be included and will be accompanied at an even later stage by a compensation system to raise extrinsic motivation. Throughout all steps the evaluation of the quality of the ratings and reviews are of concern as well. Currently, the quality of reviews is measured by the rat￾ings of reviews. No objective metric exists to measure the quality of scientific documents in an absolute way, the metric always depends on the function a document has to fulfill for a specific user. Once a reasonable number of sub￾missions of ratings of documents exists, these ratings could be compared with data from other systems like Amazon.com, data from citation indices might correlate with ratings from scientists, and an evaluation by experts (lectur￾ers, librarians, etc.) could lead to further insights as well. The most reasonable way to measure the effectiveness of the systems lies in observing the usage of the systems and asking the users, if the recommender systems (and thereby other users) helped them to find the right literature for the task they had in mind. Acknowledgments. The author gratefully acknowledge the funding of the project “Recommender Systems for Meta Library Catalogs” by the Deutsche Forschungsgemeinschaft