Speeding Up Recommender Systems Using Association Rules

Kannout, Eyad; Nguyen, Hung Son; Grzegorowski, Marek

doi:10.1007/978-3-031-21967-2_14

Cited by 4 publications

(5 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the future, we plan to investigate more algorithms for association rule and frequent pattern mining, e.g., AprioriTID or Apriori Hybrid. An important extension of the proposed method will be the introduction of additional permissions allowing for incremental processing of large data sets [40], [81] as well as enabling interaction with the user in this process of deriving a granular representation of data [25], [82]. We also plan to respond better to changes in users' behavior and preferences, addressing the possible drifts and shifts in data.…”

Section: Discussionmentioning

confidence: 99%

“…[38]) and frequent pattern mining (cf. [39]) techniques to cope with the cold-start problem is useful also when it comes to minimizing the latency of recommender systems [40]. For this reason, frequent pattern mining is particularly interesting in our research, and we provide a detailed overview of the literature on this field in Section III-B.…”

Section: Related Workmentioning

confidence: 99%

“…In addition, it provided an alternative formulation, called offsets model, that is able to make fast recommendations for new users and items and does not require any transformation for attributes data that is limited in range. On the other hand, we also evaluate our strategies against two classical and well-known latent factor models, namely Matrix Factorization (MF) [79] and Singular Value Decomposition (SVD) [5], which both rely on transforming items and users to the same latent factor or embedding space. The embedding space, also known as itemand user-embeddings, aims to explain ratings by finding the dependencies between rows (users) and columns (items) in the user-item rating matrix.…”

Section: B Baseline Recommender Systemsmentioning

confidence: 99%

“…Viable solutions to this challenge are recommender systems (RS), which leverage the data collected to provide users with the most appealing products or services they are looking for [4]. The fundamental idea of RS is to recommend items that are of interest to a user based on their previous interactions with comparable ones or categories, ratings' history of users with similar preferences, and possibly some other information, such as users' demographics or items' characteristics [5].…”

mentioning

confidence: 99%

“…Furthermore, they may be severely affected by the dynamic changes in the commercial environment where producers contend to astonish clients with ever newer and better offers. Dealing with still-emerging products and services, as well as with new visitors, is one of the most challenging tasks faced by stateof-the-art recommender systems, often referred to as the cold start problem [5], [8]. This phenomenon is particularly inconvenient and occurs whenever recommendations are generated for a new item without a proper history of user interactions or ratings.…”

mentioning

confidence: 99%

See 4 more Smart Citations

Clustering-Based Frequent Pattern Mining Framework for Solving Cold-Start Problem in Recommender Systems

Kannout,

Grzegorowski,

Grodzki

et al. 2024

IEEE Access

Self Cite

View full text Add to dashboard Cite

Recommender systems (RS) are substantial for online shopping or digital content services. However, due to some data characteristics or insufficient historical data, may encounter considerable difficulties impacting the quality of their recommendations. This study introduces the clustering-based frequent pattern mining framework for recommender systems (Clustering-based FPRS) -a novel RS constituting several recommendation strategies leveraging agglomerative clustering and FP-growth algorithms. The developed strategies combine the generated frequent itemsets with collaborative-and content-filtering methods to address the cold-start problem, which occurs whenever a new user or item enters the system. In such cases, the RS has limited information about the new user or object. Thus, the recommendations may be inaccurate. The experimental evaluation on several benchmark datasets showed that Clustering-based FPRS is superior to state-of-the-art and could effectively alleviate the cold-start problem.INDEX TERMS Cold-start problem, recommender system, frequent pattern mining, clustering I. INTRODUCTION

show abstract