%0 Journal Article %1 citeulike:201598 %A Kautz, Henry %A Selman, Bart %A Shah, Mehul %C New York, NY, USA %D 1997 %I ACM Press %J Commun. ACM %K community cscw sna %N 3 %P 63--65 %R 10.1145/245108.245123 %T Referral Web: combining social networks and collaborative filtering %U http://portal.acm.org/citation.cfm?id=245123 %V 40 %X NUMEROUS STUDIES HAVE SHOWN THAT ONE OF THE MOST EFFECtive channels for disseminating of information and expertise within an organization is its informal network of collaborators, colleagues, and friends 1, 4, 7. Indeed, the social network1 is as least as important as the official organizational structure for tasks ranging from immediate, local problem-solving (for example, fixing a piece of equipment), to primary work functions, such as creating project teams.

@article{citeulike:201598, abstract = {NUMEROUS STUDIES HAVE SHOWN THAT ONE OF THE MOST EFFECtive channels for disseminating of information and expertise within an organization is its informal network of collaborators, colleagues, and friends [1, 4, 7]. Indeed, the social network1 is as least as important as the official organizational structure for tasks ranging from immediate, local problem-solving (for example, fixing a piece of equipment), to primary work functions, such as creating project teams.}, added-at = {2006-09-25T12:54:00.000+0200}, address = {New York, NY, USA}, author = {Kautz, Henry and Selman, Bart and Shah, Mehul}, biburl = {https://www.bibsonomy.org/bibtex/2ba3606b3aa6c4cf94784db451b28cd68/grahl}, citeulike-article-id = {201598}, comment = {ReferralWeb has a number of features that sets it apart from many other collaborative filtering and recommender projects: 1. uncover existing social networds rather than providing a tool for creating new communities. 2. While recommender systems are often designed to provide anonymous recommmendations, ReferralWeb is based on providing referrals via chains of named individuals. This is critical, because not all sources of information are equally desirable. 3. Some recommender systems require the user toi manually enter a personal profile of interests, preferences, or expertise. Recommendations are generated by matching profiles that exist within the system. By constrast, ReferralWeb primarily builds its model of its users' social network by data mining public documents found on the Worls Wide Web. This model inclused many more indivisuals than those who explicitly register with the service. 4. Users of referralWebare not limited to any set of topic areas determined in advance. Referral Web uses a general full Web indexing enginve (currently, AltaVista) to match individuals to topic areas. --- Sources include: - Links found on home pages; - Lists of co-authors in technical papers and citations of papers; - Exchanges between individuals recorded in netnews archives; and - Organisation charts (e.g. for university departments)}, doi = {10.1145/245108.245123}, interhash = {6995678b936b33eef9ea1396e53a1fc7}, intrahash = {ba3606b3aa6c4cf94784db451b28cd68}, issn = {0001-0782}, journal = {Commun. ACM}, keywords = {community cscw sna}, month = {March}, number = 3, pages = {63--65}, priority = {0}, publisher = {ACM Press}, timestamp = {2007-07-25T11:36:55.000+0200}, title = {Referral Web: combining social networks and collaborative filtering}, url = {http://portal.acm.org/citation.cfm?id=245123}, volume = 40, year = 1997 }