Comparative Study on Approaches of Recommendation Systems

Fararni, Khalid Al; Aghoutane, Badraddine; Riffi, Jamal; Sabri, My Abdelouahed; Yahyaouy, Ali

doi:10.1007/978-981-15-0947-6_72

Cited by 15 publications

(20 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In contrast with item-based collaborative filtering, content-based filtering relies only on item information, not on the user’s purchase history or preferences (Al Fararni et al , 2020; Attarde and Singh, 2017). For example, when recommending a book, item-based collaborative filtering measures the similarity between books based on user preferences and then recommends books highly similar to the book preferred by the target user.…”

Section: Literature Reviewmentioning

confidence: 99%

“…Finally, a hybrid approach combines both collaborative and content-based filtering (Al Fararni et al , 2020). There are several ways to combine these two techniques.…”

Section: Literature Reviewmentioning

confidence: 99%

“…Another technique used to construct recommendation systems is called content-based filtering , which uses similarities between the contents of the items (Al Fararni et al , 2020). However, content-based filtering systems cannot use the actual contents of the books owing to copyright issues (Ng and Jung, 2019).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

What books will be your bestseller? A machine learning approach with Amazon Kindle

Lee

Kim

et al. 2021

View full text Add to dashboard Cite

Purpose With the rapid increase in internet use, most people tend to purchase books through online stores. Several such stores also provide book recommendations for buyer convenience, and both collaborative and content-based filtering approaches have been widely used for building these recommendation systems. However, both approaches have significant limitations, including cold start and data sparsity. To overcome these limitations, this study aims to investigate whether user satisfaction can be predicted based on easily accessible book descriptions. Design/methodology/approach The authors collected a large-scale Kindle Books data set containing book descriptions and ratings, and calculated whether a specific book will receive a high rating. For this purpose, several feature representation methods (bag-of-words, term frequency–inverse document frequency [TF-IDF] and Word2vec) and machine learning classifiers (logistic regression, random forest, naive Bayes and support vector machine) were used. Findings The used classifiers show substantial accuracy in predicting reader satisfaction. Among them, the random forest classifier combined with the TF-IDF feature representation method exhibited the highest accuracy at 96.09%. Originality/value This study revealed that user satisfaction can be predicted based on book descriptions and shed light on the limitations of existing recommendation systems. Further, both practical and theoretical implications have been discussed.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

“…Finally, a hybrid approach combines both collaborative and content-based filtering (Al Fararni et al , 2020). There are several ways to combine these two techniques.…”

Section: Literature Reviewmentioning

confidence: 99%

See 1 more Smart Citation

What books will be your bestseller? A machine learning approach with Amazon Kindle

Lee

Kim

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Для побудови рекомендацій такі системи використовують інформацію про продажі та рейтинги товарів, а також про зміну останніх з часом [1,3]. Однак рейтинги можуть бути спотворені в результатів шилінг-атак [4,5].…”

Section: вступunclassified

Temporal Patterns Of User Preferences In The Tasks Of Forming Explanations In The Recommender System

Чалий¹,

Лещинський²

2020

Scientific journal “Bionics of Intelligence”

View full text Add to dashboard Cite

The problem of taking into account changes in the user’s behavior of the recommendation system whenconstructing explanations for recommendations is considered. This problem occurs as a result of cyclical changes in userrequirements. Its solution is associated with the construction of an explanation comparing the alternative choices of theuser of the recommendation system. The developed models of temporal patterns consist of a set of temporal relationshipsbetween the events of users’ choice of goods and services. The first pattern contains an alternative in the form of sequential selection in time of several objects or the selection of only a pair - the first and the last object. The second pattern,sequential-alternative choice, consists of a sequence of choices over time, which ends with the first pattern. The proposedapproach to the formation of patterns is based on the construction of data sets containing temporal dependencies betweena group of user choices for a given level of time detail. The temporal dataset is used to construct a temporal graph of therecommender system user selection process. The latter includes a set of temporal patterns with an indication of the timeof their beginning and end, which makes it possible to determine the duration of the implementation of these patterns.On the basis of the patterns, subsets of temporal relationships are formed to build explanations for the recommendedlist of goods and services. Experimental verification of the developed approach using the “Online Retail” sales data sethas shown the possibility of identifying temporal patterns even on short initial samples.

show abstract

“…The existing algorithms for constructing recommendations allow you to effectively predict the interests of users in the presence of data on their purchases and ratings submitted by users [3]. However, in some situations, information supplement recommendations with explanations [6].…”

Section: Introductionmentioning

confidence: 99%

Detailing explanations in the recommender system based on matching temporal knowledge

Chalyi

Leshchynskyi

Leshchynska

2020

EEJET

View full text Add to dashboard Cite

about users may be inaccurate or incomplete, which leads to distortion of recommendations. The received incorrect recommendations do not meet the interests of users, which reduces the credibility of the recommender system. Inaccuracies and distortions in user preference data result from shilling attacks [4]. Such attacks are used to change the sales of target items in the direction desired by the attacker. The essence of a shilling attack is to artificially change the ratings of the target items. As a result, a personal recommendation is formed with distortions and may not take into account the real preferences of users. The recommender system, using an incorrect personal list of goods and services, "forces" the user to choose those items that the attacker is interested in. In case of incomplete information about the user, recommendations are built for new or irregular users [5]. In such a cold start situation, there are no data obtained as a result of feedback from the user, since the latter has not yet made purchases or posted ratings. Therefore, at present, one of the trends in the development of recommender systems, which makes it possible to consider the incompleteness and inaccuracy of the initial data, is to

show abstract

Comparative Study on Approaches of Recommendation Systems

Cited by 15 publications

References 26 publications

What books will be your bestseller? A machine learning approach with Amazon Kindle

What books will be your bestseller? A machine learning approach with Amazon Kindle

Temporal Patterns Of User Preferences In The Tasks Of Forming Explanations In The Recommender System

Detailing explanations in the recommender system based on matching temporal knowledge

Contact Info

Product

Resources

About