The Importance of Recommendation Systems in E-commerce
Recommendation systems are key to any successful e-commerce sites as they provide a better point for the buyers to make better-informed decisions as well as the business owners to get feedback and make improvements on their product performances (Kambatla et al., 2014). The current methods for collecting the user feedbacks can be used to make a better recommendation system that can be used strategically to improve the business performance in the various e-commerce campaigns. The current methods for collecting data about products reviews is unstructured and includes such techniques such as user comments, the product ratings, peer reviews among other techniques. This unstructured data cannot be analyzed by our current database structured based on structured query languages hence the need to rethink new techniques that the business can deploy to achieve a better recommendation system by exploiting the new oil for business, the information.
The report seeks to presents the market research findings from the survey done on the techniques of big data analysis. This shall be used by the board to make better decision by executing the recommendation of this report. The benefits shall not only enable the business have a competitive advantage over their rivals but also have a better insight into the business environment and performance over a long period of time. . The report next section is devoted to presenting facts and figures from our market research done on ways and techniques to implement the robust recommendation system. The last sections concluded and provide some recommendation for the board to choose from in implementing the strategies discussed in this report.
This section describes the new strategy under discussion that is the recommendation system for our e-commerce site that uses the big data to analysis. To begin with, the recommendation system is seen as a rich warehouse of information as it shall contain the user’s review of the business performance (Kitchin, 2014). When such data is analyzed using the big data techniques, the business direction shall be well predictable with a lower margin of error hence the strategic managers shall benefit from these new systems described herein making strategic decisions (Sagiroglu and Sinanc, 2013). The recommendation system provides a hybrid approach to data analysis by encompassing the following components described below;
First, collaborative-based filtering uses the user’s past purchase behavior to recommend products and services to the buyer which in most cases the user may be interested with especially if the product purchased before are complimented for the product in the filter. This cannot be made possible without the usage of big data (Hu et al., 2014).
Second, content-based filtering can also be used to recommend items to potential buyers based on their past likes and rating of items. This method is subjective and depends on the user honesty in giving their review, feedback, and ratings (Gandomi and Haider, 2015).
Apart from the hybrid approach, the recommendation system can employ the use of the conditional reviews that the customers give regarding the usage of the product and services. This particularly important for the management to fine-tune their product to meet the customers need (Davenport and Patil, 2012). With this error of information age, the customers are not only the king but also easily get access to a variety of information hence the need for the business to constantly reviewing the user feedbacks, analyzing them and given strategic plans to mitigate some of the negative reviews before they scale to catastrophic levels (Hampton et al., 2013). The following sections details how the recommendation system technique using the big data actually is within business strategic plans.
Big Data Analysis Techniques for Recommendation Systems
The new recommendation system under-discussed aligns itself in the strategic objectives of the business in the following ways;
First, it shall ensure that the business meets its strategic objectives of being number one seller of mobile devices on the e-commerce platform. This shall be made possible due to extensive algorithm envision to be implemented to ensure the recommendation systems is developed to include big data analysis tools to provide a better recommendation to our potential buyers (Lee, Kao and Yang, 2014). The usage of these new tools has made the Amazon a going concerns and currently the most profitable e-commerce venture.
Secondly, the business strategic objective to increase the share of the market by 2020 to 51% can only be achieved if the business employs the big data recommendation system and this shall provide a better way to sell more products at a lower marketing and advertising cost as the users who purchase the product shall be our ambassadors by their candid reviews and ratings of the product (Yin et al., 2013).
Furthermore, differentiation of products can only be made possible by the analysis of the unstructured user comments to gather as much information as possible so as to get better insight into the market needs of the customers which is key in making innovations to make our products much differentiated as possible hence achieving competitive advantage (Lee et al., 2013).
Last but not the least, the new big data techniques shall enable achieve its strategic objectives of improving the focus on research and development which shall enable the strategic managers to have more innovations being submitted due to the usage. Research is a core business activity hence cannot be underrated in an effort to make the business competitive (Lu et al., 2015). To ensure the success of the new strategic approach, the report has endeavored to suggest the following enabling technologies that will achieve the success in this new paradigm shift (Koren, Bell and Volinsky, 2009).
To achieve the strategies described before, the business should assign resources that are geared towards the following technologies that ensure an enabling environment for big data to ensure the recommendation system envisioned shall not only become accurate in its data interpretation but also provides an ecosystem for strategic managers to use for forecasting purposes . The ecosystem build suggested in this report requires the following technology slack.
First, the is need to invest in predictive analytics software that would use predictive algorithms to predict with a lot of accuracy the behavior and purchasing patterns for different product merchandise (Jagadish et al., 2014).
Again, Big data security solution should be employed by the management to protect this new oil in the information age. Data is so pervasive that if not protected can be used by the competitor to wreak havoc on the market share of the business. Make no mistake, cybersecurity threats are real and they’re undoubtedly have in the recent past made companies make huge loses (Siemens, 2012).
Thirdly, the business should for seeking big data solutions with in-memory data retention capabilities which makes the processes of retrieval and processing of data real time, more robust and high perform wise.
Strategic Objectives of the Business
In order to conduct the recommendation on the user data behavior and data, it is vital to report how the various analytical tools are at play to support business intelligence and decision support, the section details the working principles of the various analytical and MDM tools used in conducting the data analysis (Qian et al., 2014).
The development of the recommendation system employs the user data that the e-commerce site gathers about users. This can be explicit or implicit depending on the method employed. The explicit techniques shall use the various customer ratings and their comments on the product on the e-commerce site. Implicit data are those not directly got from the user but are rather got and analyzed using the big data tools described above to establish the most possible user behaviors concerning the product. The method that uses the behavior is relatively easier to analyze since most sites keep user log of activities they perform while on the site (Ma et al., 2015). To effectively analyze the data, the following techniques could be employed;
- Use of real-time analysis which uses the technique of processing the data as the time of creation. This can be particularly important where the management wants to know when a particular recommendation was recorded
- Use of batch analysis where a big data is collected and stored and fed into the analysis engine where much insight can be got from it. This is important in getting reports such as daily or periodic sales volume.
- Next to real-time analysis uses quick big data tools to quickly analyze the data and let the persona conducting the analysis get results by refreshing the page hosting the algorithm of the big data analysis. The algorithm is shown below,
To achieve this algorithm, there is a need to switch from the current relational database which focuses on structured data and uses the structured query language to manipulate data and perform the query. Big data, on the other hand, requires a mechanism to manipulate unstructured data hence the use of NoSQL as explained in the section below.
Master Data Management tool recommended for use is the profisee which has the ability robust data modelling of the various entities in the e-commerce site. The tool is also ideal for enumerating the business workflow for purposes of ensuring the business goals and activities are in line with the strategic objectives of the business. This enhances speedy delivery of decision for the business
The history of NoSQL can be drawn back from the internet giants such as Google, Facebook and the Amazon for their role in coming up with the NoSQL approach to analyzing big data. This has helped reduce the several drawbacks in the current RDBMS which has felt short of managing the ever-growing unstructured data sets. The discussion and support for NoSQL which has become a darling for a data scientist are discussed in this section (Sadalage and Fowler, 2013).
With the ever-growing number of Internet users which the current database system cannot handle, it became imperative for inventors to find a solution to the ever-growing big data which the online services collect about users which have NoSQL is overly important in the implementation of the recommendation system as it provides a multi-model approach to data definition, making it possible to design unstructured data sets in schema (Stonebraker, 2010).
NoSQL is flexible enough to adjust the schema dynamically as need be hence do not need to have an explicit definition done on it during application design. This is particularly important as the recommendation system shall keep some highly unstructured data such as customer ratings, web pages’ behavior, and statistics. This proves not only impossible to implement well in the traditional SQL based models but also makes the process of adjusting the schema definition time-consuming.
Conclusion
The ability of NoSQL to house several sets of databases in its data structure makes it ideal for eliminating the need to have joins operation on the data sets and the database instances are independent variables hence the NoSQL can easily scale horizontally.
To implement the recommendation system using the NoSQL approach, the following databases types was suggested to provide a more appeal to the development of the system (Li and Manoharan, 2013).
The first obvious database type that can be used to support the recommendation system is the MongoDB which is discussed in the following section
MongoDB
The MongoDB is at the center of Big Data analysis as it provided a more unstructured way that companies and corporations can use to store the data they can use to gain much insight into by querying different unstructured and unrelated datasets with the strategic objective of giving the business a competitive advantage in by better understanding the market demands.
The MongoDB stores value using the simple approach of the key-value pair where an attribute is stored with its corresponding value. The MongoDB further makes the pair stored in what they called documents which are complex data structures which can be in form of arrays, nested arrays (Parker, Poe and Vrbsky, 2013). The MongoDB is considered more relatable to the traditional RDBMS hence has achieved much acceptance among the developers, this makes it perfectly suitable for the recommendation system. Lastly, MongoDB is rich with queries that makes it perfect for the design of complex report and since it is document-oriented, it provided high flexibility, this is not only important for the complex algorithm used in the recommendation engine but also provides more strategic functionality for strategic managers for focusing and insight.
Apart from the recommendation engine, it was also envisioned from the market research that the social media plays a vital role in enhancing better decision making by the strategic manager . The findings are as shown below;
From the market research, it was evident that the business is active on the various social media with 500 000 and 300,000 thousand followers on Facebook and twitter respectively. Apparently, social media platforms are important in decision making due to the following reasons., First, it increases the scope of the organization’s collaboration and co-operation with peers and the potential customers en masse (John Walker, 2014).
Secondly, it is evident that information got from the social media has a higher trust value compared to the most survey conducted where the respondents either lie to finish the survey or the surveying team did not ask reasonable questions This has made social media the preferred platform for business surveys.
Third, with the disruption of the mobile technology, social media has even increased its volume share on the internet traffic with over 605% owned by the social media platform. This traffic when well managed can boost the company profile and sale volume by strategically investing in better ways to integrate the business key decision making by considering the surveys from the social media platforms.
References
Social media also plays a key role in rating the customer sanctification level on the business products and services.
This is key as such responses can be used by the strategic management to make some strategic plans to cater for the needs of most importantly the dissatisfied customers whose option counts as they can affect future buys.
Lastly, the social media provides a platform which can be used by the buyers to objectively evaluate the products being sold by the business hence making it form a good basis for negotiating deals.
The core importance of incorporating the big data analysis is to enhance the value chain for the business. The big data value chain is as explained below.
The value chain brought forward by the big data is as shown in the figure below,
The components represent a high-level view for the various activities that occur in the business in the scope of information systems. The components are discussed in this sections (Brown, Chui and Manyika, 2011).
This is the initial stage of capturing data and due to preliminary filters before data in input into the data marts and or warehouse. This particular component is resource intensive as it required a highly reliable system that has low latency and is capable of executing complex queries. The organization, therefore, must invest in a better system for data acquisition in the value chain.
This is the point where the data captured is modeled with a view of extracting some hidden patterns and or report which is essential for decision support. Analytical tools such as Hadoop can be used to extensively draw out the hidden patterns in the data under analysis.
Data needs to have long-term usage in a different situation. There is a constant need for the data to be remodeled to improve the quality and enhance the business access to the data by different managers. The curators ensure the trustworthiness, quality, access, and ability to discover insights .
Although the current RDBMS provides the ACID property of the database, they lack flexibility hence not suitable for big data systems such as the recommendation system which expert the schema of their database to always keep changing to meet the current need. NoSQL provides the solution by not only making the schema much flexible enough but also improves performance and scalability to tore exponential amount of data (Takeda and Kusama, 2008).
The big data has proven to be of utility to the organization in matching their recommendation engine with customer needs. This puts the business an edge by giving the management ability to act on the enormous amount of reports done from the critical analysis of the data store (Ferreira et al., 2013).
Conclusion
In this report, facts and figures have been presented to the board concerning the growing field of big data and suggested the need for the business to incorporate a recommendation system in their e-commerce site to take advantages of the various analytical tools which have been technically revised and suggestion made on the best tools set to use to get the data capture and mining possible within the business. This shall not only make the business more competitive but it will also enhance research and development of new product line which can be designed from the user comments and rating of some business products and services. The market research represented in this report has recommended the following to the board for quick actions.
The board should consider the following key points
- Set up a warehouse mart and warehouse to provide an enabling environment for big data analysis
- Invest more in the social media by employing more personnel qualified to manage the online presence of the business
- Integrate the MongoDB in the development life cycle of the business systems and application to enhance mush collaboration and machine learning among the independent systems
References
Brown, B., Chui, M. and Manyika, J., 2011. Are you ready for the era of ‘big data’. McKinsey Quarterly, 4(1), pp.24–35.
Davenport, T.H. and Patil, D.J., 2012. Data scientist. Harvard business review, 90(5), pp.70–76.
Ferreira, N., Poco, J., Vo, H.T., Freire, J. and Silva, C.T., 2013. Visual exploration of big spatio-temporal urban data: A study of new york city taxi trips. IEEE Transactions on Visualization and Computer Graphics, 19(12), pp.2149–2158.
Gandomi, A. and Haider, M., 2015. Beyond the hype: Big data concepts, methods, and analytics. International Journal of Information Management, 35(2), pp.137–144.
Hampton, S.E., Strasser, C.A., Tewksbury, J.J., Gram, W.K., Budden, A.E., Batcheller, A.L., Duke, C.S. and Porter, J.H., 2013. Big data and the future of ecology. Frontiers in Ecology and the Environment, 11(3), pp.156–162.
Hu, H., Wen, Y., Chua, T.-S. and Li, X., 2014. Toward scalable systems for big data analytics: A technology tutorial. IEEE access, 2, pp.652–687.
Jagadish, H.V., Gehrke, J., Labrinidis, A., Papakonstantinou, Y., Patel, J.M., Ramakrishnan, R. and Shahabi, C., 2014. Big data and its technical challenges. Communications of the ACM, 57(7), pp.86–94.
John Walker, S., 2014. Big data: A revolution that will transform how we live, work, and think. Taylor & Francis.
Kambatla, K., Kollias, G., Kumar, V. and Grama, A., 2014. Trends in big data analytics. Journal of Parallel and Distributed Computing, 74(7), pp.2561–2573.
Kitchin, R., 2014. Big Data, new epistemologies and paradigm shifts. Big Data & Society, 1(1), p.2053951714528481.
Koren, Y., Bell, R. and Volinsky, C., 2009. Matrix factorization techniques for recommender systems. Computer, (8), pp.30–37.
Lee, J., Kao, H.-A. and Yang, S., 2014. Service innovation and smart analytics for industry 4.0 and big data environment. Procedia Cirp, 16, pp.3–8.
Lee, J., Lapira, E., Bagheri, B. and Kao, H., 2013. Recent advances and trends in predictive manufacturing systems in big data environment. Manufacturing Letters, 1(1), pp.38–41.
Li, Y. and Manoharan, S., 2013. A performance comparison of SQL and NoSQL databases. In: Communications, computers and signal processing (PACRIM), 2013 IEEE pacific rim conference on. IEEE, pp.15–19.
Lu, J., Wu, D., Mao, M., Wang, W. and Zhang, G., 2015. Recommender system application developments: a survey. Decision Support Systems, 74, pp.12–32.
Ma, T., Zhou, J., Tang, M., Tian, Y., Al-Dhelaan, A., Al-Rodhaan, M. and Lee, S., 2015. Social network and tag sources based augmenting collaborative recommender system. IEICE transactions on Information and Systems, 98(4), pp.902–910.
Parker, Z., Poe, S. and Vrbsky, S.V., 2013. Comparing nosql mongodb to an sql db. In: Proceedings of the 51st ACM Southeast Conference. ACM, p.5.
Qian, X., Feng, H., Zhao, G. and Mei, T., 2014. Personalized recommendation combining user interest and social circle. IEEE transactions on knowledge and data engineering, 26(7), pp.1763–1777.
Sadalage, P.J. and Fowler, M., 2013. NoSQL distilled: a brief guide to the emerging world of polyglot persistence. Pearson Education.
Sagiroglu, S. and Sinanc, D., 2013. Big data: A review. In: Collaboration Technologies and Systems (CTS), 2013 International Conference on. IEEE, pp.42–47.
Siemens, G., 2012. Learning analytics: envisioning a research discipline and a domain of practice. In: Proceedings of the 2nd international conference on learning analytics and knowledge. ACM, pp.4–8.
Stonebraker, M., 2010. SQL databases v. NoSQL databases. Communications of the ACM, 53(4), pp.10–11.
Takeda, K. and Kusama, T., 2008. Data storage system. Google Patents.
Yin, H., Sun, Y., Cui, B., Hu, Z. and Chen, L., 2013. LCARS: a location-content-aware recommender system. In: Proceedings of the 19th ACM SIGKDD international conference on Knowledge discovery and data mining. ACM, pp.221–229.
Order an Essay Now & Get These Features For Free:
Turnitin Report
Formatting
Title Page
Citation
Outline
