A Framework to Specify and Implement The Back-End of S-Commerce Platforms by Micro-Services and Machine Learning

The interaction features of social web sites (social networks and social media) such as connections, communications, and content sharing are expected to enable a new kind of commerce referred to as social commerce (s-commerce) (Liang and Turban, 2012; Wang and Zhang, 2012; Zhou et al., 2013). I defined s-commerce as a new way of doing commerce in a collaborative and participative way, involving interactions among all the actors of the value chain (Baghdadi, 2016). S-commerce aims at enhancing organization openness and actor participation and collaboration to achieve greater societal and economic values. It is realized through an enterprise interactive interface that promotes social interactions. While s-commerce can be built on social web sites and e-commerce (namely B2C), it differs from both of them in many aspects, including technology platforms, system interaction, participants? connection and communication, value creation, and business model (Baghdadi 2016). For instance, e-commerce deals only with customers as individuals, whereas s-commerce sees communities of participants of different types. E-commerce is enabled by Web 1.0, whereby the communication is one-way, whereas s-commerce is enabled by Web 2.0, a platform that supports social web sites and allows bidirectional communication (for the interactions). In addition, s-commerce platforms generate a huge amount of count. This makes s-commerce platforms very challenging in terms of: (i) supporting architectures, infrastructures, platforms, and services such IoT, Big Data, Machine Learning, SOA and Cloud computing (Baghdadi, 2013), (ii) issues such as social issues, security issues, and privacy issues (Kim et al., 210), and (iii) the user perceptions, preferences, and decisions are influenced by the content generated by people interacting through social media applications (Huang and Benyoucef, 2013). While AI in general and Machine Learning (ML) in particular is applied in e-commerce in many aspects such as product recommendation, use of customer data, optimize value chain process, or make better decisions (Bawack et al. 2022). It could be extended to s-commerce. That is, realizing the back-end functionality of s-commerce platforms would require (1) a service-oriented architecture (SOA), namely micro-services and cloud computing to integrate the processes of the partner?s/suppliers architecture (Baghdadi, 2013, Baghdadi, 2016) and (2) Machine Learning (ML) for data analysis of the user generated content and recommendation models (Al Jalassi and Baghdadi, 2022). In a previous PhD project, we have worked on a framework that describes a generic set of sound requirements. These generic requirements concern with the front-end and the back-end of the platforms. That is, (1) what kind of interface the s-commerce platform should present to the participants in order to allow them to interact (participants? perspective) and (2) what kind of computation done in the back-end of a s-commerce platform (internal business?s perspective). In other works, we demonstrated how architectural pattern based on Web services and cloud (Baghdadi, 2013, Baghdadi, 2016) does support the integration with the partners and suppliers systems a required value-chain. This project concerns with the use of ML techniques and Micro-Services to specify and implement the back-end of s-commerce. More specifically, what kind of inherent computation behind the scene (back-end perspective) a s-commerce platform should provide, in terms of improving products/services/processes, via newer ML. Indeed, on one hand, ML helps in enabling any decision regarding the products/services and customer relationships to be backed by data generated by the interactions. On the other hand, micro-services interface the treatment of the user-generated content and the ML algorithms. We will conduct systematic review of primary researches, both systematic literature review (SLR) and systematic mapping study (SMS). SLR and SMS are two types of secondary research that are widely used as research methods to classify the existing literature in different domains, especially in information and software systems engineering (Kitchenham et al., 2007; Petersen et al., 2015). SLR provides an objective and unbiased approach to finding relevant primary studies by extracting, aggregating, and synthesizing the data from the primary studies (Kitchenham et al., 2007). Whereas, SMS is used to identify evidence gaps, where more primary studies are needed and evidence clusters, where it may be practical to perform a systematic review. (Petersen et al., 2015). SLR would capture the different uses of ML techniques to deal with the very huge content (e.g., Big data) generated by s-commerce platforms. The SMS will provide a map of the current applications and models of ML in s-commerce. Based on the output of the SLR and SMS, we propose to develop a framework to specify and implement s-commerce platforms using ML techniques and micro-services. The framework would provide developers of such a kind of platforms with the proper guidelines to specify and implement a full back-end s-commerce platform that is not limited only to the front-end specification. Aims of the project: The aim of the project is providing a guidance framework to specify and implement the back-end operations of a s-commerce platforms using ML techniques and micro-services. The objectives are to: - Identify the functionality of s-commerce from the back-end perspective and compare them with those of e-commerce. - Identify the ML techniques used to (analyze) generated content of a s-commerce platforms and how they can be used? - Identify the applications of ML in the development of s-commerce platforms from the back-end perspectives. - Integrate the ML within the framework of s-commerce and in which layer? - Integrate the micro-services interface the operations of s-commerce. - Conceive an architecture and a framework of s-commerce from the back-end. - Guide the developers to design s-commerce platform based on the proposed architecture/framework