Hybrid Social Recommender Systems for Electronic Commerce: A Review

Patro, S Gopal Krishna; Mishra, Brojo Kishore; Panda, Sanjaya Kumar; Kumar, Raghvendra; Apoorva, A

doi:10.1109/iccsea49143.2020.9132881

Cited by 3 publications

(5 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Social recommender systems (SRSs) can generate personalised product recommendations based on social connections between individuals (Li Y. M. et al, 2013). These systems have become an integral part of ecommerce platforms (Li Y. M. et al, 2013;Patro et al, 2020) and essential decision support systems for consumers who use such media (Chen J. & Shen, 2015;Tsai & Brusilovsky, 2021).…”

Section: Introductionmentioning

confidence: 99%

Influencer is the New Recommender: insights for Theorising Social Recommender Systems

Bawack

Bonhoure

2022

Inf Syst Front

View full text Add to dashboard Cite

Designing theory-driven social recommender systems (SRSs) has been a significant research challenge for over a decade. This study aims to identify behavioural factors that could improve the persuasiveness and quality of recommendations made by SRSs. Given both research streams' striking similarity, it uses the recent yet rich research on social media influencers (SMI) to inform SRS research. Drawing on 72 publications, we classified 52 independent variables into 12 categories regrouped into three broad categories that characterise the relationships between the consumer and the (i) recommender system, (ii) product or brand, and (iii) and advert.The metanalysis results determined the relative importance of each category in predicting purchase intentions, placing recommender credibility and attitude towards the recommended product or brand at the top of the charts.Our findings are expected to facilitate more refined theory-building efforts and theory-driven designs in SRS research and practice.

show abstract

Section: Introductionmentioning

confidence: 99%

Influencer is the New Recommender: insights for Theorising Social Recommender Systems

Bawack

Bonhoure

2022

Inf Syst Front

View full text Add to dashboard Cite

show abstract

“…However, the popularity of e-Commerce has increased the amount of information that consumers must process to be able to decide on which products/services can meet their needs. An effective solution to this information overload problem is the utilization of recommender systems [1][2][3][4][5].…”

Section: Introductionmentioning

confidence: 99%

“…The deployment of recommender systems in the e-Commerce sector has been an opportunity for constant profit and business improvement. For example, big e-Commerce companies such as Amazon, Netflix, eBay, CD-Now and Alibaba use recommender systems to promote their products and services, and enhance their customer support systems [1][2][3][4]. S c h a f e r, K o n s t a n and R i e d l [4] report that recommender systems can increase sales in e-Commerce websites by: (1) Converting the browsers who often browse e-Commerce websites without any intention of buying anything into buyers by helping them in finding the things they might like or need throughout personalized recommendations; (2) Improving the cross-selling by recommending supplementary products to buyers that might complement the products that are already chosen to buy; and (3) Improving customers loyalty by using their profiles in e-Commerce websites to offer them customized recommendations that match their needs, which increases the satisfaction of them towards services provided by such websites.…”

Section: Introductionmentioning

confidence: 99%

An Effective e-Commerce Recommender System Based on Trust and Semantic Information

Shambour

Turab

Adwan

2021

Cybernetics and Information Technologies

View full text Add to dashboard Cite

Electronic commerce has been growing gradually over the last decade as a new driver of the retail industry. In fact, the growth of e-Commerce has caused a significant rise in the number of choices of products and services offered on the Internet. This is where recommender systems come into play by providing meaningful recommendations to consumers based on their needs and interests effectively. However, recommender systems are still vulnerable to the scenarios of sparse rating data and cold start users and items. To develop an effective e-Commerce recommender system that addresses these limitations, we propose a Trust-Semantic enhanced Multi-Criteria CF (TSeMCCF) approach that exploits the trust relations and multi-criteria ratings of users, and the semantic relations of items within the CF framework to achieve effective results when sufficient rating data are not available. The experimental results have shown that the proposed approach outperforms other benchmark recommendation approaches with regard to recommendation accuracy and coverage.

show abstract

“…Tais sistemas, cada vez mais populares na Internet, tornaram-se uma alternativa para auxiliar os usuários na obtenção de itens compatíveis com o seu perfil (Bobadilla et al, 2013) (Valdiviezo-Diaz et al, 2020). Segundo Patro et al (2020), o advento dos Sistemas de Recomendação é uma fonte de lucro e aprimoramento constante no setor de comércio eletrônico.…”

Section: Introductionunclassified

“…Ambas as pesquisas, nas suas conclusões, indicaram que os métodos baseados em personalidade (personality-based) foram melhores que o tradicional (rating-based). pesquisas que não utilizaram o modelo Big Five, Srivastava, Bala & Kumar (2017) usaram o modelo Values, e os resultados indicaram desempenho melhor que a abordagem tradicional (sem uso de personalidade) Patro et al (2020). elaboraram uma proposta recente usando TKI (Thomas-Kilmann conflict mode Instrument), instrumento que funciona por meio de um questionário visando a identificar as tendências comportamentais de um indivíduo ao lidar com conflitos interpessoais; entretanto, eles ainda não avaliaram tal proposta.…”

unclassified

Estudo Comparativo de Abordagens para Sistemas de Recomendação baseados em Personalidade com uso do serviço IBM Watson Personality Insights

Aguiar¹,

Araújo²,

Costa³

2020

risti

View full text Add to dashboard Cite

Considerar a personalidade dos usuários nos Sistemas de Recomendação pode proporcionar resultados mais relevantes. Neste estudo, foi analisado se, com os avanços na detecção da personalidade (inferência sem usar questionários), as abordagens de Filtragem Colaborativa baseadas em personalidade continuam melhorando a acurácia das abordagens tradicionais (baseadas essencialmente em ratings). Outrossim, foi analisado se há diferenças ao aplicar modelos distintos de personalidade. No experimento, foram considerados 1058 usuários do TripAdvisor e 10889 clientes da Amazon, com características de personalidade inferidas via IBM Watson Personality Insights. Os resultados indicaram a possibilidade de melhorar a acurácia ao empregar uma abordagem usando dados inferidos concernentes a algum dos modelos de personalidade analisados (Big Five, Needs e Values). Ademais, o modelo Values proporcionou resultados equivalentes ao Big Five (sem facetas); e, em termos gerais, não houve melhoria ao usar o Big Five incluindo dados das suas facetas (nem ao incluir dados dos outros modelos).

show abstract

Hybrid Social Recommender Systems for Electronic Commerce: A Review

Cited by 3 publications

References 31 publications

Influencer is the New Recommender: insights for Theorising Social Recommender Systems

Influencer is the New Recommender: insights for Theorising Social Recommender Systems

An Effective e-Commerce Recommender System Based on Trust and Semantic Information

Estudo Comparativo de Abordagens para Sistemas de Recomendação baseados em Personalidade com uso do serviço IBM Watson Personality Insights

Contact Info

Product

Resources

About