Friend recommendation for healthy weight in social networks. A novel approach to weight loss

DOIhttps://doi.org/10.1108/IMDS-04-2015-0130
Date10 August 2015
Pages1251-1268
Published date10 August 2015
AuthorAnming Li,Eric W.T. Ngai,Junyi Chai
Subject MatterInformation & knowledge management,Information systems,Data management systems
Friend recommendation
for healthy weight
in social networks
A novel approach to weight loss
Anming Li and Eric W.T. Ngai
Department of Management and Marketing,
The Hong Kong Polytechnic University, Kowloon, Hong Kong,
Peoples Republic of China, and
Junyi Chai
Erasmus School of Economics, Erasmus University Rotterdam, Rotterdam,
The Netherlands
Abstract
Purpose The purpose of this paper is to propose a new approach recommending friends to social
networking users who are also using weight loss app in the context of social networks.
Design/methodology/approach Social network has been recognized as an effective way to
enhance overweight and obesity interventions in past studies. However, effective measures integrating
social network with weight loss are very limited in the healthcare area. To bridge this gap, this study
develops a measure for friend recommendation using the data obtained by weight loss apps; designs
methods to model weight-gain-related behaviors (WGRB); constructs a novel behavior network;and
develops two measurements in experiments to examine the proposed approach.
Findings The approach for friendrecommendation is based on Friend Recommendation for Health
Weight(FRHW) algorithm. By runningthis algorithm on a real data set,the experiment results show that
the algorithmcan recommend a friend who has ahealthy lifestyle to a target user.The advantages of the
proposedmechanism have been well justifiedvia comparisons with popular friendrecommenders in past
studies.
Originality/value The conventional methods for friend recommenders in social networks are only
concerned with similarities of pairs rather than interactions between people. The system cannot account
for the potential influences among people. The method pioneers to model a WGRB as recommendation
mechanismthat allow recommended friendsto simultaneouslyfulfill two criteria. They are:first, similarity
to the target person; and second, ensuring the positive influence toward weight loss. The second criterion
is obviously important in practice and thus the approach is valuable to the literature.
Keywords Healthcare, Obesity, Social network, Recommendation
Paper type Research paper
1. Introduction
1.1 Background
Recommendation system (RS) processes the information by suggesting to users the
objects that are possibly compatible with their interests (Adomavicius and Tuzhlin,
2005; Kim, 2011). RS has been comprehensively investigated in different perspectives
such as preference (Koutrika and Ioannidis, 2010), trust (Bedi et al., 2007), decision
making (Chai et al., 2014), and position localization (Bao et al., 2015). Online social Industrial Management & Data
Systems
Vol. 115 No. 7, 2015
pp. 1251-1268
©Emerald Group Publis hing Limited
0263-5577
DOI 10.1108/IMDS-04-2015-0130
Received 14 April 2015
Revised 31 May 2015
Accepted 2 June 2015
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0263-5577.htm
The authors are grateful for the constructive comments of the referees on an earlier version of
this paper. The third author thanks LEI Chun-Fong for helpful comments.
1251
Novel
approach to
weight loss
networks offer users opportunities to expand their circle of friends through a RS
(Curras-Perezet al., 2014). The system breaks regionalrestrictions and unites like-minded
people who have similar interests or goals (Yang et al.,2012).Forexample,asoccerfan
may associate with other fans to enjoy FIFA World Cup.
Online social networks are great sources of health-related information and social
support. They provide access to people who share the same experience and 24/7
encouragement, which can be difficult for realistic friends. Goel and Goldstein (2014)
reported the application of social networks to predict the behavior of individuals with
respect to the large social data in marketing environments. Centola (2010) investigated
the influence of social network structure on diffusion by studying the spread of health
behavior through artificially structured onlinecommunities.Hefoundthattheproliferation
of behavior across clustered-lattice networks would be farther and faster than that across
corresponding random networks.
A weight management program could be a long-term process and requires constant
self-monitoring. Individuals are likely to concede in the early stage and eventually fail
the weight management program. An obese user who participates in a weight management
program may want to know other participants in the program in social networks, thus
promoting each other for better achievement. Weight loss apps respond to such requests
and typically require users to monitor their food intake and physical activities to calculate
calorie consumption.
1.2 Literature review
Existing literature provided some valuable solutions for better weight management.
Maher et al. (2014) reviewed the effectiveness of interventions based on social network.
Those interventions aimed to change health behaviors such as dietary intake and
physical activity. Among the studies, 90 percent reported positive outcomes.
Chang et al. (2013) reviewed the role of social media in online weight management.
They pointed out that social media could help those who lack social support establish
connections and obtain support. Hwang et al. (2010) investigated peer support in an
online weight loss community, and suggested that 60 percent of the participants
considered online members were more helpful than other contacts on the issu e of
weight loss. Convenience, anonymity, and none-judgment are the three distinct features
valued most by the respondents.
Li et al. (2013) explored the role of social networks and social media in obesity
intervention. They identified three pathways that link obesity with social networks,
namely, social support, social integration, and social capital. They advocated for the
development of social networks specifically designed to address obesity. However,
Kiernan et al. (2012) and Gorin et al. (2005) found that obesity intervention involving
spouse, friends, co-workers, and neighbors had mixed results. Kiernan et al. (2012)
found that women with family support were more likely to lose weight (71.6 percent),
whereas women who did not obtain support from friends were most likely to lose
weight (80 percent). Gorin et al. (2005) demonstrated that participants with one
or more successful partners lost more weight compared with those without successful
partners.
A landmark study by Christakis and Fowler (2007) used longitudinal statistical
models to examine whether weight gain is contagious. They showed that the risk of a
person becoming obese increased by 57 percent if a friend became obese. They further
revealed that the involvement of partners in weight management process could provide
support and emotional buffer to people who face a similar problem. In other words,
1252
IMDS
115,7

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT