Adaptive E-commerce System - Literature review Example

1030986 - ADAPTIVE E-COMMERCE SYSTEM To say that the coming of technology and the advent of the World Wide Web and the internet has made modern business more lucrative is a fact that none can deny. As technology leads the pace on globalisation and join the whole together into a common platform where the exchange of information across geographic borders have been made so easy, businesses continue to have a fair share of the situation. According to quote (year), one of the ways that companies have greatly benefited from technology is through the concept of electronic commerce (e-commerce). Throughout literature, there continue to be some levels of differences in the definition of e-commerce. Most of these differences arise with differing framework of what e-commerce constitutes. For example quote (year) looked at e-commerce from the perspective of buying and selling and stated that e-commerce covers all industries where the processes of buying and selling is performed over electronic systems. Quote (year) on the other hand viewed e-commerce from the perspective of what is done with a company’s website. To this end, all activities that promoted the regulation of business, whether it involved selling and buying or not formed e-commerce. Examples of such activities were named to include optimization of traffic from search engines, generation of sales and interaction with customers through an online platform. Regardless of the differences in framework for defining e-commerce however, there is one thing that runs through and that is the fact that some level of business that takes place over an electronic medium. As part of the purpose of the study, the literature review is being conducted as a major secondary data collection exercise that will be used to collect data to address the specific objectives. From the background of the study, it would be noted that there is great regard for adaptive e-commerce. This means that the researcher shall approach the literature from a perspective that limits the scope of review to adaptive e-commerce system. This is because of the position held by the researcher that even though e-commerce as a generalised concept of modern business has the potential of opening great doors for companies, there is the need to optimise the outputs of e-commerce. By optimisation, reference is being made to any creative process that ensures that service users are able to have an advantage on the use of e-commerce over their competitors. Meanwhile, quote (year) made claim to the effect that with the competitiveness of the modern global economic environment, it is important that business operators have a means to personalizing their target market and customers. This way, they will be able to offer tailored customised and user based services and products to customers. By extension, customers will prefer to do business with such service providers that give customised service because they will be noted to understand the needs of customers better (quote). It is the need to serving this purpose of personalising one’s market that the adaptive e-commerce is being reviewed to ensure that through that platform, e-commerce could be personalised and adapted to suit specific needs of companies. 2.1 Agent-Based e-Commerce Architecture 2.11 Adaptation and Personalisation Throughout literature, the call for adaptation and personalisation seem to focus on two major area of discussion. The first has to do with the school of thought that adaptation and personalization must be approached from the perspective of the host or service provider. Debating for stand, quote (year) stressed the fact that in modern day e-commerce, customers expect that they will be served. For this reason, customers expect surprises from their service providers. Once customers would use a particular e-commerce platform, they want to see rapid changes that serve their needs. Because of this, service providers are expected to personalize their e-commerce platforms in such a way that meet the unique ways in which the service providers can best provide service according to the dynamic expectation of their market. Quote (year) supported this argument and stated that one of the best ways to ensure a service provider based adaptation and personalisation is to ensure that there is rigorous market research that identifies expectations of customers. Writing from the other side of the debate, quote (year) mentioned that customers are now the drivers of the global economy and so instead of expecting to satisfy the customer by surprising them, it is better to satisfy them by making them choose how they should be served. It is not surprising that major companies such as Tesco have taken up the idea and allowed personalisation of their e-commerce in a way that makes it possible for the customer to dictate the pace of service (quote). Adding to thought to this school of thought of customer based personalisation, quote (year) noted that traditional stores are seen to have some level of advantage over online stores because with traditional stores, customers are able to have a personal feel of the shopping experience. The solution to overturning this ideology is by making sure that whiles using e-commerce platforms, customers can also have a say of serve they want that meet their needs (quote). But fairly argued, each side of the debate presents some level of merit, that calls for a hybrid adaptation and personalization approach to e-commerce (quote). 2.12 Adaptive Hypermedia Modelling For the most parts of the birth of e-commerce, a major criticism that has gone for traditional e-learning and e-government systems is that they offer too many links, making it difficult for users to choose from and get their preferred service needs met (quote). The introduction of adaptive hypermedia is therefore seen by quote (year) as a solution to this situation where users get lost in hyperspace. This is because quote (year) mentioned that with adaptive hypermedia, links and contents that are peculiar to the needs of users are provided. Regardless of the general solution that adaptive hypermedia offers to users, quote (year) noted that it is not all times that the needs of users are totally served, given that user needs can be highly varying. As a way of further personalizing adaptive hypermedia platforms, the use of various adaptive hypermedia models is recommended. Three of such adaptive hypermedia models mentioned by quote (year) include user model, domain model and interaction model. Quote (year) has explained the user adaptive hypermedia model as the type that makes use knowledge source in making explicit assumptions about the user. For such systems therefore, inferences are made through the usage pattern the user from their periodic interaction with the system. The domain model on the other hand focuses on applications that need to be present to make the operation of the adaptive system possible. The applications of a typical adaptive hypermedia would therefore be seen through different phases and levels including task level, logical level and physical level (quote). Touching on the interaction model, quote (year) explained that this model of the adaptive hypermedia serves as a midpoint between the user model and domain model. This is because it defines the relationship that ought to exist between these two to ensure that the system is operating within the paradigms of an adaptive hypermedia. Quote (year) stressed on the interdependence nature of the three models and mentioned that the full capacity of an adaptive hypermedia will be hampered if each level is not seen to be functioning as it ought to. 2.13 Adaptive Hypermedia layers Using the user model and adaptive models of adaptive hypermedia, The University of Warwick (2009) identified three major adaptive hypermedia layers and the different roles that they play in a typical adaptive model. Together, these three layers form the layered adaptation granulation (LAG). The first layer is referred to as the lowest lever, which is basically responsible for generating direct adaptation techniques and rules (University of Warwick, 2009). According to quote (year), at the lowest level, all that has to do with the regulation of the adaptive model is developed because that is the point that functionality is built by way of the adaptation rules that control the functioning of the model. This idea is supported by the University of Warwick (2009) as the lowest level is further referred to as the adaptation assemble language. By implication, the notation of the adaptive hypermedia is created by way of the output of the lowest level of the layer. Quote (year) also emphasised on the lowest layer, saying that to start off well, it is important to employ all needed programming tools that can adequately write programs that relate various variables to data. Even though there may be the use of programming tools at the lowest level, the actual output by way of programming takes place at the next level, which is the medium level, also referred to the adaptation language layer (quote). Quote (year) stated that at the medium level, it is expected that the direct adaptation rules on language will be implemented into actual adaptation language. This language is said by quote (year) to be the adaptation programming language that introduces new techniques that were absent at the lowest level. Quote (year) however warned about the nature of language at this layer and how this should be differentiated from language at the lowest level by stating the importance of importing higher level language right at this stage. At the third level is the high level, which is the adaptation strategies layer. At this point, the need to input all adaption function calls is emphasised (quote). It is for this reason that Warwick (2009) mentions this point to be goal-oriented and a wrapping of the above layers. Figure 1: Layered Adaptation Granulation (LAG) Source: Warwick (2009, p. 8) 2.2 Domain Models of Adaptive e-Commerce Already, mention has been made of the domain model as one of the key architectures for the adaptive system. In this section of the review, specific cases of domain models in existing systems are critically reviewed to understand their role in adaptive systems for e-commerce. 2.21 Domain Models in the ADAPTS Project The ADAPTS project represents the Adaptive Diagnostics and Personalised Technical Support project, which provides an adaptive electronic performance support system (EPSS) for maintaining complex equipment (Brusilovsky and Cooper, 2002). Within the ADAPTS is domain models that can best serve the purpose of replica domain models for various adaptive architecture. Describing the domain models in the ADAPTS, Brusilovsky and Cooper (2002) explained that the basic functionality of the domain it to provide intelligence performance key to the system. This is done purposely to provide troubleshoot for maintenance tasks to be performed (quote). Within the ADAPTS therefore, the domain model “mirrors the hierarchy of systems, subsystems, and components in the target system” (Brusilovsky and Cooper, 2002, p. 3). Like in the adaptive hypermedia layers described above the domain model of the ADAPTS is also said to have three layers, which are the lowest, medium and high levels. Each of the layers is however given a unique coding that relates to its function as shown in the figure below. Figure 2: Structure of the ADAPTS Domain Model Source: Brusilovsky and Cooper, 2002, p. 3 The diagram above depicts an ADAPTS domain model that is functioning in a larger aircraft system. The diagram explains the three major layers, the first of which is the lowest level, represented by the addressable unit. The addressable unit refers to components of the system that are subject to manipulation by the technicians who will work on the larger system (quote). The addressable unit has often been limited in terms of user knowledge as it operates based on the knowledge where it is found alone. The second level, which is the medium layer, is represented as the replaceable unit. As the name implies, the replaceable unit of the domain refers to the parts that can be removed and reinstalled at any point in time by the technician working on the system. Over here, experience is versatile and may either be limited to the location or open to the user (quote). The highest level is represented by the system or subsystem, which is made up of a lot of sub-components as depicted in the diagram. Brusilovsky and Cooper (2002) therefore noted that there is a theory of operation related to the highest layer that is measured by a user’s knowledge about the system. 2.22 Domain Models in MOT MOT is denoted by My Online Teacher, which is seen more as an educational adaptive hypermedia that is used to deliver educational courses (quote). Having been successfully developed and extended at the Eindhoven University of Technology, the MOT is considered by many who are concerned with domain models in e-commerce as an ideal mould that can be studied for its domain model functionality for replication in similar adaptive hypermedia (quote).g given that a typical MOT has four major modules in its architecture, each of which is divided into sub-components, quote (year) noted that the module that serves the purpose of domain model is the adaptation module. This deduction is made based on the functionality and the scope of the adaptation module of the MOT. From the diagram below, it would be noted that the adaptation module is made up of two major components, just as the other components that also has different numbers of components. Figure 3: Modules and Components in MOT 2.0 Source: Ghali, Cristea and Stewart (2009). The two components of the adaptation module are Adaptive Module and Adaptive Strategy. Like a typical domain model, quote (year) noted that these two components come together to determine the features of application which can be adapted for the system. This way, the three itinerant roles of task level, logical level and physical level are all played in their coded formats (quote). In the MOT, the adaptive module serves part of this role as it uses static modules and adaptive parameters to determine outcome of user module attribute component. More specifically, Ghali, Cristea and Stewart (2009) explain that the static module contains different adaptive items such as text, image, video and audio. These items further have attributes stored as key-value pairs, which meet users preferences, skills and knowledge-level to enact usability (Ghali, Cristea and Stewart, 2009). For the adaptive strategy, they are more focused on the logic level as they are used to define rules about static modules using clues from the related adaptive parameters and user modules (quote). 2.23 Domain Model in ADE Adaptive display environment (ADE) is another adaptive system that has served as a basis for determining the domain model, ideal to function in a typical adaptive hypermedia. Like most other forms of authoring adaptive hypermedia tools, quote (year) noted that the ADE is a system that is used to display personalised and adapted content for any specific user. In this system, the basic functionality of personalisation and adaption is made to revolve around five major parameters. At each point within these parameters, a specific modal role is performed. But the point within the five parameters that focuses on applications necessary for the operation of the entire adaptive system is the adaptation structure (quote). Through the adaptation structure, which acts as the domain model for ADE, it is possible to dramatically increase a variety of adaptation specifications (Scotton, Stewart and Cristea, 2009). Quote (year) added that it is for this singular reason that it is possible to that the ADE can be developed on embedded devices for digital signage. More specifically on the adaptation structure, Scotton, Stewart and Cristea (2009) explained that it contains adaptation specifications that are similar to adaptive strategies in MOT. These adaptation specifications in ADE, unlike the MOT are stored independently from the content as a means of ensuring that there is maximum reusability (Scotton, Stewart and Cristea, 2009). Because of the reusability component of the ADE, it is always possible and permissible to apply on several courses (quote). This means that the adaptation structure is in place to specifically ensure that the ADE is adaptation language independent, forming and utilising its own adaptation language (quote). Coupled with the fact that the domain model in the form of the adaptation structure uses a modular system for adaptation, ADE is able to provide an interpreter module, based on the adaptation language that originates from the system. Indeed, due to such multi functional applicability of the ADE, Scotton, Stewart and Cristea (2009) considers it to have a better navigation facility, speed and layout and usability than most other forms of systems including AHA! 2.3 Adaptation Strategy Electronic advertisement and for that matter online advertisement is becoming a very popular part of e-commerce and the general use of the internet. Figures given from the Interactive Advertising Bureau (2013) indicated that each year from 2011, there is a climb of 18% in revenues from e-advertisement. This figure resulted in US 7.8 billion in 2011 and USD 9.26 billion in 2012. Even though these figures may sound very impressive, quote (year) opined that given the size of internet users and the rate at which internet user base increase, the 18% climb in revenues from e-advertisement is not the best that can be realised. This assertion was made in comparison to the 24% increase in internet activity by the year (quote). In effect, quote (year) said that e-advertisement faces a major challenge that has to do with its effectiveness and popularity. Quote (year) stated that the problem with e-advertisement is that it may be popular and cost advertisers a lot of billions of dollars but the actual question should be about the returns that advertisers make out of the investment they make from e-advertisement. In relation to the larger discussion on e-commerce, it can be noted that it is not enough to have a platform that promotes e-commerce but that this platform must result in revenue generation for the people who invest in them. It is against the backdrop of the fact that most internet users do not respond positively to e-advertisement that the need for a strategy to ensure that internet users follow advertisement and make a positive impact out of them is necessary (quote). Quote (year) mentioned that even though there has been the use of a number of strategies to make e-advertisement effective, the use of adaptive e-advertisement is the most popular and one that has yielded the most of results among advertisers. Adaptive strategy to e-advertisement is explained as a form of e-commerce strategy where advertisers ensure that users of the internet are made to attach personal and social domains to online advertisements (quote). Quote (year) noted that the use of adaptive strategy in e-advertisement is most common among social networking companies and other companies that major in e-advertisement. With this, names such as Facebook, Amazon and Google easily come up. The adaptive strategy used by these companies has been explained by quote (year) to be the one that attaches personal and social domain to e-advertisement by ensuring that there is a means by which personal information from users are acquired. In most of these platforms, users will be presented with various applications and programmes, which they may want to explore. These applications and programmes that appear on their devices are not the actual versions of what they want to use. By clicking to use the application, a user is taken to another page, where the need to sign up is required. By signing up, the user is made to share important personal information, which is retained and used by the social networking host (quote). Because companies require user behaviour and characteristics to make important decision on marketing, they would also purchase such adapted information. The adaptive strategy used by these social networking sites to extract user information has thus been compared to a platform that promotes the presentation of personalised advertising (quote). The adaptive strategy has been explained to have a 2-way benefit, where it helps both the advertisement host such as Google and company wanting to market their outfit. A typical example of this platform can be cited using Google’s AdSense. Through Google AdSense, an adaptive approach is taken to undertaking interactive media advertisement targeted at particular site content and audience (quote). For Google Network as the host of the adaptive strategy, it has to use various means to get as many user information based on user behaviour and characteristics as possible. Once this is done, Google gets the right audience for each type of advertisement. Some of the means by which a network host like Google, Facebook, Amazon or any other would use to generate user information is through various sign up information that users exchange (quote). With the assurance that the right audience is secured for each form of product or service, Google would now place AdSense advertisements on various platforms that they have such as Google+ or Gmail. As users click on these advertisements, Google makes revenue for the pay-per-click service offered. Even though Google, Facebook, Amazon and the likes may be bigger names in adaptive e-advertisement, quote (year) noted that every other company has the opportunity of making the most out of their websites through the use of various adaptive strategies. For these individual companies, all they need to do is to have the right targeting approach to ensure that users will not only view advertisements but will at least click on them to gain information. 2.4 Authoring Systems and Tools When a person opens any website, there are two major things that the person may be offered from the website. These are receptive web surfing or productive web surfing (quote). Receptive web surfing has been explained as that form of web service that the user or visitor is only allowed to receive what is being offered by the website without having the means to contribute to the website (quote). The productive web surfing on the other hand has been explained as that active interactive web experience where the user or visitor can actually manipulate the website and thus become a productive user of the website (year). Limiting the discussion to e-commerce, quote (year) noted that in a modern age of e-commerce, productive use of websites is the first approach to successful e-commerce. But to achieve a productive state for one’s website, there are some important jinxes that must be broken. According to quote (year), authoring is the best means to produce a productive website via the use of authoring systems. To this end, authoring systems have been explained to be programmes with pre-programmed features that make it possible for users to develop multimedia applications for manipulating multimedia objects (quote). On their part, the W3C (2014) explained that authoring tools are said to reach their peak of usability when they allow all users “to publish to a universal space of web content, read by people from all over the world, in many different languages, on many computers, using many different input and output devices” (W3C, 2014). From the complexity of the demand output expected of authoring, one can expect that authoring will not be achieved on just any other platform. It is for this reason that there are specific authoring systems and tools that are used to serve the purpose of authoring. According to quote (year), there are several authoring tools with each tool playing a well defined role. Web page authoring tools such as WYSIWYG HTML editors are examples of authoring tools that are used to create interactive websites (quote). There are also software for directly editing source code or markup and software for converting to web content technologies such as “Save as HTML” (W3C, 2014). Other authoring tools are software that is used to rapidly update portions of web pages ushc as blogging, wikis and online forums (W3C, 2014). Other authoring tools are also designed to play the specific role of allowing live collaboration over the web. Quote (year) mentioned that even though authoring tools have been used over the years to serve several purposes, including the creation of social media websites such as Twitter and online trade sites where the transaction of sales take place such as eBay, the commonest forms of authoring tools are used for educational purposes. This means that most forms of authoring tools are directed at the creation of e-learning modules. In order to ensure the uniformity of such modules, they are often designed to comply with international standards, of which names like Shareable Content Object Reference Model (SCORM) can be mentioned. Quote (year) has also named My Online Teacher (MOT) as a popular model authoring system that has transitioned through several stages of authoring developments and upgrading. Today, several researchers have attempted design upgrades for MOT to ensure that the system is well netted into the demands of modern e-commerce. For example, Ghali, Cristea and Stewart (2009) focused on a MOT system that has a collaborative authoring and social annotation between communities of authors and also ensuring greater level of adaptive control to user activities. As an authoring system, MOT 2.0, which is an upgraded version of MOT, is equipped to make suggestions on related materials and recommend collaborators, based on an author’s interaction patterns Ghali, Cristea and Stewart (2009). An outstanding feature of the MOT 2.0 as an authoring system has been explained by quote (year) to be the system’s ability to harness Web 2.0 and there guarantee group-based adaptation. With a group-based adaptation, the emphasis on the system’s authoring abilities become shifted from the act of singling out specific authors based on their interaction pattern into the process of coordinating the activities of several authors at the same time. Up to date therefore, the MOT has been described as one of the most resilient authoring systems with adaptive hypermedia-based authoring for personalization (quote). With its adaptive hypermedia-based authoring, authors from different backgrounds are able to have a social annotation that is based on the use of multimedia course materials that can be manipulated and directed at the achievement of specific study goals. 700 2.5 Web Services, 2.51 Productive and Receptive Web Services Since the birth of e-commerce, the number of web services that are offered on the internet have been noted to shoot up to very high levels. Today, the comparative growth rate in the addition of new services on an annual rate is pegged at 46% by quote (year). Whereas some of these web services are receptive in nature and hardly allow any interactivity and manipulation by users, others have taken the productive approach through authoring platforms as have been outlined above. The major line of argument that has been seen in existing literature therefore has to do with the merits and demerits of each form of web service, be it productive or receptive. On the use of productive web services, quote (year) noted that in an era of user empowerment and adaptation, one can expect that the best answer to people based web service is by the use of productive web services. This is because these productive web services that are mounted on the platforms of authoring systems ensure that users have maximum control over the services they are using and are able to be a social part of these services. On the part of quote (year), it was stated that the real case of argument must be about the purpose that a particular web service is being used to offer that should determine whether productive or receptive approach will be the best. For example when it comes to services such as social media networking, there cannot be any doubt that using a productive platform that allows users to interact with the website and thus their social connections is the best choice. Writing about receptive web services, quote (year) pointed out that there are times that service hosts are left with no other choice than using this form of web service. This is especially when it comes to web services that have to do with financial issues such as the purchasing of products online. Even though users may be allowed some form of interaction through emails and messaging platforms, allowing absolution productive activities may expose the website to activities of hackers and hijackers, as productive web services have been noted to be generally weak in security issues (quote). Using receptive web services are therefore a means of enhancing the privacy and security of websites. The debate also gets very interesting in terms of academic based web services. This is because whereas quote (year) noted that productive educational web services such as wikis and blogs makes it possible to gain universal contribution to educational topics and issues, quote (year) wondered if this did not expose such web service to questions of quality. This is because in most of the cases, the identities and credibility of contributors are not known and thus expose the service to quality issues. As a solution to the use of productive web services for educational purposes, quote (year) advised that there can be strategies and platforms for checking the authenticity of contents before putting them up. Because this will be a whole lot of work for the website host, the ambition of the hosts must not be about the quantity of content they have but the quality of these (quote). 2.52 Service-Oriented Architecture Whether by the use of a productive or receptive web service, service providers have often had the need for service-oriented architecture to make their services activated. As shown in the diagram below, Quote (year) explained that service-oriented architecture (SOA) is a pattern in software design and software architecture that uses discrete pieces of software to enable application functionality as services to other applications. In other words, SOA is a platform for providing application functionality service to other applications in the web service venture (quote). For this reason, there is often the service provider and service consumer, all of who rely on the service broker. In the context of the argument therefore, the service broker becomes the software architecture design pattern. The term service orientation is used to denote the fact that this form of architecture is independent of an specific vendor or technology, meaning that it allows a self-contained functionality for whoever wants to make use of a particular service platform. References University of Warwick (2009) Authoring of Adaptive Hypermedia. Accessed March 17, 2014 from http://www.dcs.warwick.ac.uk/~acristea/Books/AHChenBookChapt-cameraReady2.pdf Brusilovsky and Cooper (2002). Domain, Task, and User Models for an Adaptive Hypermedia Performance Support System. ACM. Vol. 3 No. 2; pp 2-8 Ghali F., Cristea A.I and Stewart C (2009). My Online Teacher 2.0. Coventry: University of Warwick GoogleAdSense, "Maximize revenue from your online content" www.google.com/adsense/ [accessed 25 March 2013]. Scotton J, Stewart C. and Cristea A. I (2009). ADE: The Adaptive Display Environment for Adaptive Hypermedia. Accessed March 17, 2014 from http://www.ht2011.org/demos_posters/ht2011_submission_144.pdf InternetAdvertisingBureau. (2012). Internet Advertising Revenues Hit Historic High in Q3 2012 at Nearly $9.3 Billion, Rising 18% Over the Same Period Last Year, According to IAB. Accessed March 15, 2014 from http://www.iab.net/about_the_iab/recent_press_releases/%20press_release_archive/press_release/pr-121912 W3C (2014). Authoring Tools, Social Media. Accessed March 19, 2014 from http://www.w3.org/standards/agents/authoring Ghali F, Cristea A.I and Stewart C (2009). My Online Teacher 2.0. Coventry: University of Warwick, Read More