Wyniki wyszukiwania - Biblioteka Nauki

1

Selection of clusters based on internal indices in multi-clustering collaborative filtering recommender system

100%

Kużelewska U.

|

2024

|

tom Vol. 70, No. 1

103--114

EN

The successful application of a multi-clusteringbased neighborhood approach to recommender systems has led to increased recommendation accuracy and the elimination of divergence related to differences in clustering methods traditionally used. The Multi-Clustering Collaborative Filtering algorithm was developed to achieve this, as described in the author’s previous papers. However, utilizing multiple clusters poses challenges regarding memory consumption and scalability. Not all partitionings are equally advantageous, making selecting clusters for the recommender system’s input crucial without compromising recommendation accuracy. This article presents a solution for selecting clustering schemes based on internal indices evaluation. This method can be employed for preparing input data in collaborative filtering recommender systems. The study’s results confirm the positive impact of scheme selection on the overall recommendation performance, as it typically improves after the selection process. Furthermore, a smaller number of clustering schemes used as input for the recommender system enhances scalability and reduces memory consumption. The findings are compared with baseline recommenders’ outcomes to validate the effectiveness of the proposed approach.

2

The recommendation algorithm for an online art gallery

100%

Karwowski W. , Sosnowska J. , Rusek M.

|

tom Vol. 7, No. 2

108--119

EN

The paper discusses the need for recommendations and the basic recommendation systems and algorithms. In the second part the design and implementation of the recommender system for online art gallery (photos, drawings, and paintings) is presented. The designed customized recommendation algorithm is based on collaborative filtering technique using the similarity between objects, improved by information from user profile. At the end conclusions of performed algorithm are formulated.

3

Holistic Entropy Reduction for Collaborative Filtering

75%

Szwabe A. , Misiorek P. , Janasiewicz T. , Walkowiak P.

|

tom Vol. 39, No. 3

209--229

EN

We propose a collaborative filtering (CF) method that uses behavioral data provided as propositions having the RDF-compliant form of (user X, likes, item Y ) triples. The method involves the application of a novel self-configuration technique for the generation of vector-space representations optimized from the information-theoretic perspective. The method, referred to as Holistic Probabilistic Modus Ponendo Ponens (HPMPP), enables reasoning about the likelihood of unknown facts. The proposed vector-space graph representation model is based on the probabilistic apparatus of quantum Information Retrieval and on the compatibility of all operators representing subjects, predicates, objects and facts. The dual graph-vector representation of the available propositional data enables the entropy-reducing transformation and supports the compositionality of mutually compatible representations. As shown in the experiments presented in the paper, the compositionality of the vector-space representations allows an HPMPP-based recommendation system to identify which of the unknown facts having the triple form (user X, likes, item Y ) are the most likely to be true in a way that is both effective and, in contrast to methods proposed so far, fully automatic.

4

Stereotype-Aaware Collaborative Filtering

75%

Frisch G. , Leger J.-B. , Grandvalet Y.

|

2021

|

tom Vol. 25

69--79

EN

In collaborative filtering, recommendations are made using user feedback on a few products. In this paper, we show that even if sensitive attributes are not used to fit the models, a disparate impact may nevertheless affect recommendations. We propose a definition of fairness for the recommender system that expresses that the ranking of items should be independent of sensitive attribute. We design a co-clustering of users and items that processes exogenous sensitive attributes to remove their influence to return fair recommendations. We prove that our model ensures approximately fair recommendations provided that the classification of users approximately respects statistical parity.

5

Products and services recommendation systems in e-commerce. Recommendation methods, algorithms, and measures of their effectiveness

63%

Lasek M. , Kosieradzki D.

|

nr 1(31)

304-317

EN

The article concerns products and services recommendation systems in ecommerce which have become increasingly important for both consumers and retailers. The methods used for the recommendation of products and services, as well as the algorithms used to implement them, are presented in the article. Particular attention was paid to the problems of testing the suitability of algorithms, along with the effectiveness measures of the applications of the methods and algorithms.

6

Context Clustering-based Recommender Systems

51%

Kannout E.

|

tom Vol. 21

85--91

EN

Recommender systems have gained lots of attention due to the rapid increase in the amount of data on the internet. Therefore, the demand for finding more advanced techniques to generate more useful recommendations becomes an urgent. The increasing need for generating more relevant recommendations led to the emergence of many novel recommendation systems, such as Context-aware Recommender System (CARS), which is based on incorporating the contextual information in recommendation systems. The goal of this paper is to propose new recommender systems that utilize the contextual information to find more relevant recommendations.