Feature Fusion Recommendation Algorithm Based on Collaborative Filtering

Wang, Kexin; Chen, Zhanfang; Wang, Yuanshuai; Yang, Zhangnan

doi:10.1109/mlbdbi48998.2019.00040

Cited by 4 publications

(4 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On the other hand, due to the different consumption behaviors and habits of the consumers in the different regions of a city, some information on the behaviors might get lost or be ignored, leading to false recommendations. In that light, Wang et al [ 85 ] proposed a feature fusion personalized recommendation algorithm based on collaborative filtering combined with dense feature data and sparse feature data. This algorithm focused on learning the characteristics of users in a specific region and the characteristics of some sparse users while it also learned the time period of ordering, thus mitigating the model and personalizing the recommendation of the item the consumer may buy.…”

Section: Resultsmentioning

confidence: 99%

Applications of Fusion Techniques in E-Commerce Environments: A Literature Review

Daskalakis¹,

Remoundou²,

Πεππές³

et al. 2022

Sensors

View full text Add to dashboard Cite

The extreme rise of the Internet of Things and the increasing access of people to web applications have led to the expanding use of diverse e-commerce solutions, which was even more obvious during the COVID-19 pandemic. Large amounts of heterogeneous data from multiple sources reside in e-commerce environments and are often characterized by data source inaccuracy and unreliability. In this regard, various fusion techniques can play a crucial role in addressing such challenges and are extensively used in numerous e-commerce applications. This paper’s goal is to conduct an academic literature review of prominent fusion-based solutions that can assist in tackling the everyday challenges the e-commerce environments face as well as in their needs to make more accurate and better business decisions. For categorizing the solutions, a novel 4-fold categorization approach is introduced including product-related, economy-related, business-related, and consumer-related solutions, followed by relevant subcategorizations, based on the wide variety of challenges faced by e-commerce. Results from the 65 fusion-related solutions included in the paper show a great variety of different fusion applications, focusing on the fusion of already existing models and algorithms as well as the existence of a large number of different machine learning techniques focusing on the same e-commerce-related challenge.

show abstract

Section: Resultsmentioning

confidence: 99%

Applications of Fusion Techniques in E-Commerce Environments: A Literature Review

Daskalakis¹,

Remoundou²,

Πεππές³

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…However, this two hyper-parameter cannot be updating constantly， thus it has extremely weak generalization ability to predict the error from missing rating [10]. According to this weakness, double learning has been proposed which utilized two separate matrix factorization structures and different parameters to predict imputation model and rating model respectively [7,11]. Moreover, during the training of one model, another model will also be updated simultaneously.…”

Section: Joint Learning and Data Fusionmentioning

confidence: 99%

“…Also, the EIB estimator has difficulty estimating the estimation error accurately, while the IPS estimator often suffers from the problem of large variance, i.e., when the estimated tendency is very small, it creates a very large value. Hence the appearance of the Doubly Robust estimator [5,7], which integrates estimated error and propensity in a doubly robust manner to obtain unbiased performance estimates and mitigate the effects of propensity variance. That is, as long as the estimated propensity is accurate, or the estimated error is accurate.…”

Section: Introductionmentioning

confidence: 99%

Causal Fusion for Recommender System

Sun

Zhang

Zi-lin

et al. 2023

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

Recommender system suffers from huge selection bias from users, which makes utilizing causal inference to solve this problem becomes a necessary problem. However, traditional rating model only utilizing biased dataset and unbiased estimator cannot remarkably be improved. Thus, we fused the unbiased dataset into training model and combine it with jointly learning method to better improve the original baselines. Extensive experiments verify that the proposed doubly robust joint leaning method after fusing unbiased, missing rating data can significantly outperform the state-of-the-art methods and can dramatically reduce the error from rating model with the increasing number of unbiased data.

show abstract

“…Density can be computed as d = #Ratings #Users * #Items . To ascertain that a single rating value range was used, in order for the results of the different datasets to be comparable, the ratings in each dataset were normalised in the range [1.0-5.0], using the standard min-max formula [42], which is used in many CF research works [43][44][45]. In order to quantify the rating prediction accuracy, the following two CF rating prediction error metrics were used [46,47]:…”

mentioning

confidence: 99%

On Exploiting Rating Prediction Accuracy Features in Dense Collaborative Filtering Datasets

2022

View full text Add to dashboard Cite

One of the typical goals of collaborative filtering algorithms is to produce rating predictions with values very close to what real users would give to an item. Afterward, the items having the largest rating prediction values will be recommended to the users by the recommender system. Collaborative filtering algorithms can be applied to both sparse and dense datasets, and each of these dataset categories involves different kinds of risks. As far as the dense collaborative filtering datasets are concerned, where the rating prediction coverage is, most of the time, very high, we usually face large rating prediction times, issues concerning the selection of a user’s near neighbours, etc. Although collaborative filtering algorithms usually achieve better results when applied to dense datasets, there is still room for improvement, since in many cases, the rating prediction error is relatively high, which leads to unsuccessful recommendations and hence to recommender system unreliability. In this work, we explore rating prediction accuracy features, although in a broader context, in dense collaborative filtering datasets. We conduct an extensive evaluation, using dense datasets, widely used in collaborative filtering research, in order to find the associations between these features and the rating prediction accuracy.

show abstract

Feature Fusion Recommendation Algorithm Based on Collaborative Filtering

Cited by 4 publications

References 1 publication

Applications of Fusion Techniques in E-Commerce Environments: A Literature Review

Applications of Fusion Techniques in E-Commerce Environments: A Literature Review

Causal Fusion for Recommender System

On Exploiting Rating Prediction Accuracy Features in Dense Collaborative Filtering Datasets

Contact Info

Product

Resources

About