Purpose – Online social networks have become one of the main feeder of news and it has been demonstrated their ability to determine opinion formation of users (Xiong et al., 2017). The growing number and consequences of fake news are capturing attention and concerns of the biggest players of the market such as Facebook and Google (The New York Times, 2016) who risk to lose credibility (The Washington Post, 2016). Another source of concern is the publication of inappropriate news (due to topic or multimedia contents not compliant with users’ community). This paper proposes an innovative method to cope with fake and inappropriate news on social networks based on an extension of the Grouber model belonging to the class of Collective Knowlegde Systems (CKS) which incorporates a Consensus Method (Herrera-Viedma et al., 2014) typically used in the Group Decision Making (Alonso et al., 2013) built according to the Viable System Approach (VSA) perspective (Barile and Polese, 2010) and taking into account engagement (Carrubbo et al., 2017). Design/Methodology/approach – Social networks are assumed as a specific example of Collective Knowlegde Systems (CKS), and an example of them is the model of Grouber (Grouber, 2008) in which small groups of proactive users produce information artifacts that can be searched by other users which need information. Using the VSA lenses (Barile et al. 2015), these groups will be identified for having a specific information variety that allows them solving the complexity of news credibility/appropriateness assessment in a timely way. Given the Collective Intelligence logic (Levy, 1994), these people would need heterogeneous knowledge (but not dissonant information varieties), multiplying instead of summing their intelligence (Kerckhove, 1996) in order to provide different micro-contributions to the understanding (Nielsen, 2012). This would dynamically change their information variety and the information variety of the system as a whole (the social network). Moreover, the model will represent an engagement platform in the social network (which is, according to Storbacka et al. (2016) the microfundation of value co-creation) of both the judging experts ( reworded with a competence score) and the other users (who will perceive the news reliability in a consonant relation with the community). By adopting an holistic and systems perspective, such system’s reaction (thanks to its autopoietic traits) to reset relations with actors of the system’s structure in order to re-configure the system towards a new equilibrium can be interpreted as the system tension to keep viability. After an introduction on the issue (section 1) and a comprehensive literature review related to service system and VSA on one side and CKS, Grouber model and consensus methods on the other side (section 2), the model will be mathematically formalized and described in detail (section 3) and finally shown in a simple application. Social and managerial implications will close. Findings – The method proposed in the paper answers to the following questions: 1) How can fake/inappropriate news be detected by groups of individuals belonging to an on-line social network? 2) How can these individuals be further engaged in the community? The method could be efficiently adopted in private social networks within organizations but also extended to bigger and general purpose ones. Research limitations/implications - This research moves on the new frontier of social networks autopoiesis and will provide novel interesting contributions to the reflections on decision making in systems and viability of Smart Service Systems (Polese et al., 2017). Further researches will be needed to find out the optimal (and minimum) number and composition (in terms of information variety and gained rewording score) of experts to assign to each news to be evaluated in the consensus phase. Moreover, semiautomatic semantic filters could be introduced in the model to both classify news within expert classification categories and reduce the total amount of news to be evaluated (and, consequently, the lead time needed to make decisions), finally making the methods operating in a “quasi-real time” social network environment. Originality/value – The originality of the paper relies on: • adoption of the VSA lenses to cope with the issue of fake/inappropriate news in on-line social networks; • adaptation of the Grouber model to the issue and incorporation of a consensus method; • introduction of a further engagement logic into the community.
A new consensus method for social networks viability
Sarno D;
2017-01-01
Abstract
Purpose – Online social networks have become one of the main feeder of news and it has been demonstrated their ability to determine opinion formation of users (Xiong et al., 2017). The growing number and consequences of fake news are capturing attention and concerns of the biggest players of the market such as Facebook and Google (The New York Times, 2016) who risk to lose credibility (The Washington Post, 2016). Another source of concern is the publication of inappropriate news (due to topic or multimedia contents not compliant with users’ community). This paper proposes an innovative method to cope with fake and inappropriate news on social networks based on an extension of the Grouber model belonging to the class of Collective Knowlegde Systems (CKS) which incorporates a Consensus Method (Herrera-Viedma et al., 2014) typically used in the Group Decision Making (Alonso et al., 2013) built according to the Viable System Approach (VSA) perspective (Barile and Polese, 2010) and taking into account engagement (Carrubbo et al., 2017). Design/Methodology/approach – Social networks are assumed as a specific example of Collective Knowlegde Systems (CKS), and an example of them is the model of Grouber (Grouber, 2008) in which small groups of proactive users produce information artifacts that can be searched by other users which need information. Using the VSA lenses (Barile et al. 2015), these groups will be identified for having a specific information variety that allows them solving the complexity of news credibility/appropriateness assessment in a timely way. Given the Collective Intelligence logic (Levy, 1994), these people would need heterogeneous knowledge (but not dissonant information varieties), multiplying instead of summing their intelligence (Kerckhove, 1996) in order to provide different micro-contributions to the understanding (Nielsen, 2012). This would dynamically change their information variety and the information variety of the system as a whole (the social network). Moreover, the model will represent an engagement platform in the social network (which is, according to Storbacka et al. (2016) the microfundation of value co-creation) of both the judging experts ( reworded with a competence score) and the other users (who will perceive the news reliability in a consonant relation with the community). By adopting an holistic and systems perspective, such system’s reaction (thanks to its autopoietic traits) to reset relations with actors of the system’s structure in order to re-configure the system towards a new equilibrium can be interpreted as the system tension to keep viability. After an introduction on the issue (section 1) and a comprehensive literature review related to service system and VSA on one side and CKS, Grouber model and consensus methods on the other side (section 2), the model will be mathematically formalized and described in detail (section 3) and finally shown in a simple application. Social and managerial implications will close. Findings – The method proposed in the paper answers to the following questions: 1) How can fake/inappropriate news be detected by groups of individuals belonging to an on-line social network? 2) How can these individuals be further engaged in the community? The method could be efficiently adopted in private social networks within organizations but also extended to bigger and general purpose ones. Research limitations/implications - This research moves on the new frontier of social networks autopoiesis and will provide novel interesting contributions to the reflections on decision making in systems and viability of Smart Service Systems (Polese et al., 2017). Further researches will be needed to find out the optimal (and minimum) number and composition (in terms of information variety and gained rewording score) of experts to assign to each news to be evaluated in the consensus phase. Moreover, semiautomatic semantic filters could be introduced in the model to both classify news within expert classification categories and reduce the total amount of news to be evaluated (and, consequently, the lead time needed to make decisions), finally making the methods operating in a “quasi-real time” social network environment. Originality/value – The originality of the paper relies on: • adoption of the VSA lenses to cope with the issue of fake/inappropriate news in on-line social networks; • adaptation of the Grouber model to the issue and incorporation of a consensus method; • introduction of a further engagement logic into the community.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.