Recommender Systems: A Vital Tool for Web Users - Essay Sample

Introduction

In a world of the web that people live in today, recommender systems have become vital tools as people seek a recommendation for the best items. Mansur et al. (2019) and Chen et al. (2019) discuss in great depth the recommender systems that web users could use when looking for items. A recommender system is one that filters information in a bid to predict the preference of a user. Mansur et al. (2019) discuss the three main categories of recommender systems that usually used depending on the information they apply. The three recommender systems categories are; collaborative, content-based, and hybrid filtering. Whereas Mansur et al. (2019) focus on the three recommender systems categories, Chen et al. (2019) focus on addressing the challenges involved with the collaborative filtering recommendation, from which they recommend a new model.

Is your time best spent reading someone else’s essay? Get a 100% original essay FROM A CERTIFIED WRITER!

Save 25%

Both articles appreciate the power of information. However, the need for recommender systems has been amplified by the overload of information online. Recommender systems help filter the information and tailor suggestions to the needs or desires of the specific user. The collaborative filtering approach features prominently in both papers. Chen et al. (2019) discuss the evolutionary clustering algorithm as used in this approach. Other algorithms are similarity indices and the UCEC&CF Algorithm Flow. The similarity indices algorithms for suggestions and recommendations are also discussed in context by Mansur et al. (2019).

Chen et al. (2019) discuss user correlation indices. In this algorithm, a rating is derived from an average of reviews by different users. It is done depending on their experience after purchase. On the other hand, Mansur et al. (2019) have an approach they call the user-based collaborative filtering. They have a more complex algorithmic formula, but it points to the same thing as the user correlation model discussed by Chen et al. (2019). The main concern from both discussions is the reliability and accuracy of the algorithms, due to the uniqueness of every single user.

Mansur et al. (2019) observed some of the most common challenges across all the algorithms. The challenges include difficulty trying to recommend to new users, little trust in the conclusions made from little data, user data privacy, and sparsity of data. They advise that one has to rise above the challenges to use a recommender system effectively. If one is still in the shadows of the challenges, the recommendations may not be in line with user needs and preferences.

Chen et al. (2019) go to great depth to develop a new algorithm for a collaborative recommender approach; they name it User Correlation and Evolutionary Clustering algorithm (UCEC&CF algorithm). Their experiments and results lead to the CF model's recommendation, which is a hybrid of the evolutionary clustering and the user correlation algorithms. From the extensive experimental data obtained in the research, Chen et al. (2019) conclude that the new collaborative filtering model achieves a higher recommendation performance than the other available algorithms. It is in itself a progressive achievement in the area of recommender systems.

The research and conclusions by Mansur et al. (2019) get amplified by the acts of Chen et al. (2019). The former recommended more research to ensure that a more effective and better performing algorithm comes into use. The new algorithm would put into consideration the shortcomings of the existing ones. The authors agree that recommender algorithms will become even more critical in the future, hence the need to ensure they are of great help to web users. As it stands, Chen et al. (2019) have tried to improve the existing collaborative filtering system. Next, researchers could improve the hybrid and content-based filtering approaches, making them more effective than they are currently.

Related Work and Background on Block-Chain Recommendation Systems

An increased amount of information on the web has made it fundamental for e-commerce sites to use recommender systems to make suggestions to users (Lisi et al., 2019). The recommender systems have been established to help a consumer make a relevant decision, buy using already available data to make recommendations. In this approach, a blockchain refers to a system of data structure holding records of transactions, under no specific jurisdiction or authority. This data can be used to create a rating system that users then apply when selecting new products to purchase. Lisi et al. (2019) claim the recommender system is managed and controlled centrally. Global preference and relevance are the main pointers when making suggestions to clients. The provided ratings influence the decision of the purchaser.

Lisi et al. (2019) recommend the smart contracts approach of recommender systems. In this case, the smart contract prevents modification of the main blocks, and the peer-to-peer review gets replicated across the network. The blockchain technology is highly prevalent in the cryptocurrency industry, and it can be replicated in other areas like e-commerce. In the background information about blockchain technology, Lisi et al. (2019) refer to the Ethereum platform. It is the most common open-source system that features the smart contract element.

The smart contract function in blockchain technology uses the rating system to place some products and brands higher in the ranking than others. The natural; path f selection is that consumers will select the product or brand with the best rating and ignore the rest. The most common algorithm in this approach is the Proof of Work (PoW) algorithm. Apart from an in-depth discussion about the smart contract approach, Lisi et al. (2019) discuss other aspects of recommender systems. The authors explain the wide-ranging application of recommender systems in different platforms online, opining that the usage can only get bigger and broader.

When YouTube recommends a video, or Amazon recommends a product, or Facebook suggests an article; they have applied the smart contract recommender system; depending on a user's past history. Lisi et al. (2019) feel that the recommendation or suggestion must not be trivial, and very critical evaluation and rankling system must be used. There must be a great deal; of personalization when suggesting if optimum results are to be achieved. After all assessment of what is available, Lisi et al. (2019) recommend a system that incorporates smart contracts and recommender system to give a user the best experience. In such a scenario, they believe that the needs of the user shall be put into primary consideration before those of the product maker.

Lisi et al. (2019) use both the Ethereum and Ropsten platforms to conduct their tests and make a proper recommendation. After the study and practical research, the authors recommend a framework of a smart contract that is externally compatible with a recommender system. The proposed approach should be context-free. Such should help bring on board transparency, decentralization, and immutability. In this framework, the users should have a simple access mechanism whenever they want to rate a product. The new position results from the current situation where a product rating is seen as off-the-chain. The proposed system should help prevent an attack on massive negative reviews that are usually really helpful in broadening user information before making a purchase decision. The freedom to do a review rests with a user; the product owner does not have the control to overwrite such a review.

References

Chen, J., Zhao, C., Uliji, & Chen, L. (2019). Collaborative filtering recommendation algorithm based on user correlation and evolutionary clustering. Complex & Intelligent Systems, 6(1), 147-156. https://doi.org/10.1007/s40747-019-00123-5

Lisi, A., De Salve, A., Mori, P., & Ricci, L. (2019, September). A Smart Contract Based Recommender System. In International Conference on the Economics of Grids, Clouds, Systems, and Services (pp. 29-42). Springer, Cham.

Mansur, F., Patel, V., & Patel, M. (2019). A review on recommender systems. 2017 International Conference On Innovations In Information, Embedded And Communication Systems (ICIIECS). https://doi.org/10.1109/iciiecs.2017.8276182