Research on Personalized Recommendation Methods for Online Video Learning Resources

E-commerce recommender system based on improved K-means commodity information management model

Since the start of the 21st century, there has been a rapid development of internet technology, causing electronic computers and smartphones to become increasingly popular. The e-commerce industry also experiences quick development. However, the recommendation technology of e-commerce progresses slowly, hindering it from keeping up with the changing times. To enhance the efficiency and accuracy of e-commerce recommender systems, this research introduces an e-commerce recommender system that utilizes an enhanced K-means clustering algorithm to manage commodity information. This method combines the K-means algorithm with a genetic algorithm by encoding the genetic algorithm, setting the initial population, defining the fitness function, and configuring other parameters. The results of the test indicated that the K-mean clustering algorithm and fuzzy C-mean algorithm had a recommendation accuracy of 87.9 % and 84.8 % respectively under the test dataset. The highest recommendation accuracy was observed from the improved K-mean clustering algorithm, which was 91.1 %. The convergence rate of the improved K-mean clustering algorithm was faster by 44 % compared to the traditional K-mean clustering algorithm and 73 % quicker than the fuzzy C-mean algorithm. The study's findings demonstrate that the refined K-means clustering algorithm greatly enhances the recommendation proficiency and precision of the e-commerce recommendation system, in comparison to other comparable algorithms. This research can potentially advance the e-commerce industry and stimulate its growth.

"Guess You Like It" - How Personalized Recommendation Timing and Product Type Influence Consumers' Acceptance: An ERP Study

Personalized recommendation has been increasingly used in online shopping environment, and improving the effectiveness of personalized recommendation is an important issue. On the basis of two-stage decision theory and preference inconsistency theory, our study adopted the neuroscientific methodology of event-related potential to investigate the decision-making process and psychological mechanism of consumers for personalized recommendation under different recommendation timings (browsing and decision stages) and recommended product types (similar and related). Behavioral results showed that consumers' acceptance of similar product recommendations was higher than that of related product recommendations during the browsing stage, whereas no difference was observed in consumers' acceptance of the two product types during the decision stage. More importantly, neurophysiology results provided underlying psychological mechanism for exploring consumers' decision-making process for personalized recommendations. Consumers' psychological mechanism of the personalized recommendations might be divided into two processes, the early automatic cognitive process indicated by the N2 component, and the late advanced cognitive process indicated by the P3 component. We suggested that N2 reflects the perceptual mismatch between the recommended products and the target products, and P3 reflects the attention capture during categorization evaluation of the recommended product and the target product. These findings have important theoretical and practical significance for the deeper understanding of consumers' decision-making process and psychological mechanism in personalized recommendation, as well as improving the effectiveness of personalized recommendation.

Association Rules Mining for Hospital Readmission: A Case Study

As an indicator of healthcare quality and performance, hospital readmission incurs major costs for healthcare systems worldwide. Understanding the relationships between readmission factors, such as input features and readmission length, is challenging following intricate hospital readmission procedures. This study discovered the significant correlation between potential readmission factors (threshold of various settings for readmission length) and basic demographic variables. Association rule mining (ARM), particularly the Apriori algorithm, was utilised to extract the hidden input variable patterns and relationships among admitted patients by generating supervised learning rules. The mined rules were categorised into two outcomes to comprehend readmission data; (i) the rules associated with various readmission length and (ii) several expert-validated variables related to basic demographics (gender, race, and age group). The extracted rules proved useful to facilitate decision-making and resource preparation to minimise patient readmission.