Application Of Recommender Systems Research Articles

A graph neural approach for group recommendation system based on pairwise preferences

Pairwise preference information, which involves users expressing their preferences by comparing items, plays a crucial role in decision-making and has recently found application in recommendation systems. In this study, we introduce GcPp, a clustering algorithm that leverages pairwise preference data to generate recommendations for user groups. Initially, we construct individual graphs for each user based on their pairwise preferences and utilize a graph convolutional network to predict similarities between all pairs of graphs. These predicted similarity scores form the foundation of our research. We then construct a new graph where users are nodes and the edges are weighted according to the predicted similarities. Finally, we perform clustering on the graph’s nodes (users). By evaluating various metrics, we found that employing a similarity metric based on a convolutional neural network (SimGNN) with our proposed ground truth called Top-K yielded the highest accuracy. The proposed approach is specifically designed for group recommendation systems and holds significant potential for group decision-making problems. Code is available at https://github.com/RozaAbolghasemi/Group_Recommendation_Syatem_GcPp_clustering.

Application of Makeup Image Optimization Recommendation System through the Analysis of BeautyGAN Based on Deep Learning

The purpose of this study was to identify the makeup preference of users and suggest a method to optimize the makeup style by using the preferred image for each age group through the analysis of BeautyGAN. Through this, you can propose a customized makeup style that suits you, and provide beneficial services to the makeup industry and consumers. In addition, by developing and validating new methods that effectively combine deep learning and vision systems, we aim to innovate makeup-related image conversion technology and contribute to academic and practical advances in this field. For this purpose, reference images suitable for each image were collected to implement image optimization for each age group, the input data reflected the researcher’s image, and the face was aligned and resized, after removing images with low resolution or poor lighting conditions. As a result of the performance evaluation of the BeautyGAN model, it was confirmed that the existing image was 51.26%, which is close to the BeautyGAN image of 38.89%. These results are judged to be able to provide customized makeup style suggestions or adjusted makeup effects that reflect the user’s preferences from an academic point of view, and from a practical point of view, it will be possible to improve the quality of customized beauty services by suggesting makeup styles that suit the characteristics of customers more accurately and quickly.

GSRec: A Graph-Sequence Recommendation System Based on Reverse-Order Graph and User Embedding

At present, sequence-based models have various applications in recommendation systems; these models recommend the interested items of the user according to the user’s behavioral sequence. However, sequence-based models have a limitation of length. When the length of the user’s behavioral sequence exceeds the limitation of the model, the model cannot take advantage of the complete behavioral sequence of the user and cannot know the user’s holistic interests. The accuracy of the model then goes down. Meanwhile, sequence-based models only pay attention to the sequential signals of the data but do not pay attention to the spatial signals of the data, which will also affect the model’s accuracy. This paper proposes a graph sequence-based model called GSRec that combines Graph Convolutional Network (GCN) and Transformer to solve these problems. In the GCN part we designed a reverse-order graph, and in the Transformer part we introduced the user embedding. The reverse-order graph and the user embedding can make the combination of GCN and Transformer more efficient. Experiments on six datasets show that GSRec outperforms the current state-of-the-art (SOTA) models.

Designing a Dynamic Weighted Stacking Recommendation System

To increase accuracy, ensemble approaches have become increasingly popular in designing recommendation systems. More recently, efforts have focused on dynamically integrating base models to improve ensemble performance further. Dynamic integration entails combining multiple base models with dynamically changing contributions to the ensemble. In this paper, we propose a Dynamic-Weighted Stacking (DWS) recommendation model. Bayesian optimization is used to adjust hyperparameters and find the best weight for each base model dynamically. The base models used in the DWS recommendation model are K-Nearest Neighbors (KNN), Linear Regression (LR), and Support Vector Machine (SVM). The proposed ensemble model is tested across various scenarios corresponding to different user populations with varying ratings and user counts, and the results are reported. We also compare the DWS recommendation model to individual static base models. It is observed that the batch dynamic integration approach works better than individual static models and static traditional stacking. The proposed model's flexibility in handling various data distributions highlights its potential for real-world applications in recommendation systems.

Evaluation of Machine Learning Techniques for Motivational Quotes Classification and Categorization

The effective classification and categorization of motivational quotes are critical for various applications in content management and recommendation systems. This paper presents a comprehensive methodology for consolidating and categorizing motivational quotes using machine learning techniques. The dataset, sourced from Kaggle’s Quotes-500K, contains over half a million quotes categorized into various tags such as ”Love,” ”Life,” ”Motivation,” and more. We address category redundancies by consolidating similar tags and selecting appropriate embedding models to capture semantic relationships between quotes. The embedding models are evaluated using t-SNE for finding the most distinctive model for representing motivational quotes. The final categorization focuses on three key categories: ”Love,” ”Inspirational,” and ”Humor”. We then evaluate the performance of several AI models, including Neural Networks, K-Nearest Neighbors (KNN), Euclidean distancebased classifiers, and Cosine Similarity methods. The proposed improved neural network model outperformed other models with a test accuracy of approximately 98%. These results suggest that our approach offers a robust solution for scalable and accurate categorization of motivational content. Future work will explore deeper NLP techniques, fine-tuning strategies for not just increased number of categories but further enhancing the categorization accuracy along with ranking them within their category.

Implementation and Application of Employment Recommendation System for College Students with Personalized Preferences

Abstract With the popularization of higher education, the competition in the employment market of college students is becoming increasingly intense. To enhance the employment efficiency and satisfaction of college students. The study first analyzes college students’ employment unit selection, attribute preference, and location preference through an employment recommendation algorithm. The collaborative filtering algorithm is utilized to complete personalized modeling and output the final recommendation results based on the acquired employment preferences and relevant data collection. Finally, the gradient descent method is used to evaluate the accuracy of college students’ employment recommendations. The results show that the overall educational requirements of enterprises for the three significant positions of short video production, account operation and anchor are not high, and the percentage of those with education of high school or below or master’s degree or above is meager, neither exceeding 3.5%. The personalized employment recommendation algorithm can provide scientific and reasonable guidance for graduates’ employment, with an accuracy rate of up to 50% and a recommendation list length of N=30. When α=0.75, the personalized employment recommendation algorithm can obtain better recommendation performance with smaller recommendation list length. This paper provides new solutions for college students’ employment and valuable references and lessons for research in related fields.

A one-step graph clustering method on heterogeneous graphs via variational graph embedding

<abstract><p>Graph clustering is one of the fundamental tasks in graph data mining, with significant applications in social networks and recommendation systems. Traditional methods for clustering heterogeneous graphs typically involve obtaining node representations as a preliminary step, followed by the application of clustering algorithms to achieve the final clustering results. However, this two-step approach leads to a disconnection between the optimization of node representation and the clustering process, making it challenging to achieve optimal results. In this paper, we propose a graph clustering approach specifically designed for heterogeneous graphs that unifies the optimization of node representation and the clustering process for nodes in a heterogeneous graph. We assume that the relationships between different meta-paths in the heterogeneous graph are mutually independent. By maximizing the joint probability of meta-paths and nodes, we derive the optimization objective through variational methods. Finally, we employ backpropagation and reparameterization techniques to optimize this objective and thereby achieve the desired clustering results. Experiments conducted on multiple real heterogeneous datasets demonstrate that the proposed method is competitive with existing methods.</p></abstract>

Consumer Behaviour: Analysing Marketing Campaigns through Recommender Systems and Statistical Techniques

This article examines consumer behaviour’s impact on marketing campaigns' effectiveness using a recommender system and statistical analysis methods. Understanding consumer behaviour is essential in today's fiercely competitive and constantly evolving market. Our study aims to highlight the significant impact of consumer behaviour on marketing data through the innovative application of recommender systems supported by state-of-the-art machine learning and data analysis techniques. This approach addresses the formidable challenges of accurately predicting consumer behaviour. We provide a detailed introduction to recommendation systems, emphasizing their vital role in the modern marketing landscape. We then outline our theories, laying the groundwork for a deeper understanding of the relationship between marketing data and consumer behaviour. Additionally, we present a rigorous data analysis process that begins with data cleaning and progresses through univariate and bivariate analysis, culminating in advanced techniques such as the Apriori algorithm to discover association rules and thoroughly explore this symbiotic relationship. Our findings demonstrate the applicability and effectiveness of our methodology for interpreting the complex interplay between consumer behaviour and marketing data. Our conclusions highlight essential trends and offer practical recommendations for enhancing marketing strategies significantly. By elucidating the dynamic relationships between consumer behaviour and marketing outcomes, our study contributes to a more sophisticated understanding of consumer dynamics in the contemporary business environment. Furthermore, this paper underscores the importance of understanding consumer behaviour and the benefits of employing innovative data analysis methods. By decoding consumption trends, businesses can optimize their marketing strategies and improve customer satisfaction, strengthening their competitive edge in a constantly shifting market. Finally, incorporating recommender systems with artificial intelligence and machine learning tools for collaborative filtering can further refine these strategies, substantially boosting marketing efficacy.

Based on intelligent advertising recommendation and abnormal advertising monitoring system in the field of machine learning

With the rapid development of the Internet, the scale of the online advertising market has expanded rapidly, and display advertising has become the most popular means of publicity. Accurate advertising recommendation is the guarantee of Internet platform revenue, and accurate advertising click rate prediction is the premise of accurate recommendation and abnormal advertising detection. Therefore, such monitoring and recommendation can be achieved through machine learning combined with artificial intelligence, and the application of intelligent AD recommendation systems and abnormal AD monitoring in the field of machine learning represents a complex integration of technologies to improve the precision and effectiveness of digital marketing strategies. Intelligent AD recommendation systems utilize advanced machine learning algorithms to analyze user behavior and preferences to deliver tailored AD content. These systems leverage vast amounts of user data, including browsing history, purchase history, and engagement metrics, to predict and present the most relevant ads. This paper analyzes the data mining in machine learning algorithms and the real-time online recommendation algorithm of Gaussian process, and analyzes the abnormal advertising monitoring system for maintaining the integrity and efficiency of advertising campaigns. By using machine learning technology for pattern recognition and anomaly detection, various measures and indicators of advertising campaigns can be monitored in a vigilant manner.

Music emotion classification using a hybrid CNN-LSTM model

The emotional content of music, interwoven with the intricacies of human affect, poses a unique challenge for computational recognition and classification. With the digitalization of music libraries expanding exponentially, there is a pressing need for precise, automated tools capable of navigating and categorizing vast musical repositories based on emotional contexts. This study advances music emotion classification in the field of music information retrieval by developing a deep learning model that accurately predicts emotional categories in music. The goal of this research is to advance the field of music emotion classification by leveraging the capabilities of convolutional neural networks combined with long short-term memory within deep learning frameworks. The contribution of this study is to provide a refined approach to music emotion classification, combining the power of convolutional neural networks and long short-term memory architectures with sophisticated preprocessing of the Emotify dataset for a deeper and more accurate analysis of musical emotions. The research introduces a novel architecture combining Convolutional Neural Networks and Long Short-Term Memory networks designed to capture the intricate emotional nuances in music. The model leverages convolutional neural networks for robust feature detection and Long Short-Term Memory networks for effective sequence learning, addressing the temporal dynamics of musical features. Utilizing the Emotify dataset, comprising tracks annotated with nine emotional features, the study expands the dataset by segmenting each track into 20 parts, thereby enriching the variety of emotional expressions. Techniques like the synthetic minority oversampling technique were implemented to counter dataset imbalance, ensuring equitable representation of various emotions. The spectral characteristics of the samples were analyzed using the Fast Fourier Transform, contributing to a more comprehensive understanding of the data. Through meticulous fine-tuning, including dropout implementation to prevent overfitting and learning rate adjustments, the developed model achieved a notable accuracy of 94.7 %. This high level of precision underscores the model's potential for application in digital music services, recommendation systems, and music therapy. Future enhancements to this music emotion classification system include expanding the dataset and refining the model architecture for even more nuanced emotional analysis

Social networks data analytical approaches for trust‐based recommender systems: A systematic literature review

SummaryWith the explosive growth of information on the web and the quick provision of novel web services, recommendation systems have emerged as efficient mechanism for providing advice regarding items in which users might be potentially interested. However, a traditional recommender system, which solely mines the user's previous behavior and item descriptions for recommendations, has some drawbacks. To overcome these limitations, a new solution that has recently garnered significant attention is using trust data. This approach, however, presents challenges in utilizing suitable trust data according to recommender systems applications, underlying social network structures, and user needs. In addition, in a selective decision‐making system, trust, as a kind of social network data, plays a significant role and needs an appropriate approach for making inferences. This article provides a systematic literature review of the current trust‐based social recommender systems. It also presents a detailed categorization of the trust type utilized and inferred from the existing trust‐related social networks data analytical approaches. Furthermore, it addresses the main properties and challenges of the most popular trust‐based social recommendation systems. Finally, it presents our findings and discusses open issues that provide researchers with insight to develop more enhanced recommender systems.

Doubly constrained offline reinforcement learning for learning path recommendation

Learning path recommendation refers to the application of interactive recommendation systems in the field of education, aimed at optimizing learning outcomes while minimizing the workload of learners, teachers, and curriculum designers. Reinforcement Learning (RL) has proven effective in capturing and modeling the complex interactions among course activities, learner behaviors, and educational outcomes. Therefore, combining the two approaches presents endless possibilities for personalized education through the use of interactive recommendation systems in the education domain. However, traditional RL algorithms require extensive interaction with the environment during the training phase. Using unverified recommendation logic in interactions with actual students can give rise to unmanageable problems and hinder effective performance in an educational setting. This is because extrapolation introduces substantial evaluation errors that result in recommendations deviating significantly from the actual educational requirements. To address this limitation, we propose a novel method of offline reinforcement learning called Doubly Constrained deep Q-learning Network (DCQN). This method utilizes two generative models to fit existing student historical interaction data, which in turn, constrains the original policy network to generate new actions based on past interactions, avoiding the occurrence of overestimated actions and reducing extrapolation errors. Empirical results on demonstrate that this approach performs better than existing techniques across D4RL, i.e., datasets for deep data-driven reinforcement learning and real educational datasets.

Realization and Application of Knowledge Sharing Recommendation System for Power Grid Enterprises

Abstract Facing the complex knowledge of power grid enterprises, this paper builds and designs a system for sharing project knowledge. The design pattern of MVC is adopted as the business logic architecture of the system, and a large-scale distributed application based on JavaEE architecture is used. In the knowledge recommendation method, a recommendation algorithm based on knowledge context is proposed, and the similarity between different entities is calculated by matching the knowledge ontology, thus realizing the recommendation of enterprise knowledge based on entity similarity. Meanwhile, the recommendation effect is analyzed by stress testing the recommendation service interface as a performance test characterization of the system and combined with the hit rate. The hit rate of recommendation based on knowledge context is 0.124 higher than that of BPRMF, 0.104 higher than that of NeuMF, and 0.083 higher than that of CKE. This system is able to effectively improve the knowledge-sharing efficiency of power grid enterprises.

Machine Learning based Movie Recommendation System

Abstract: Machine Learning (ML) could be a fashionable engineering technique to form machines suppose or use their intelligence like humans by mimicking traits and by learning to require acceptable choices and to perform appointed tasks properly. a number of the businesses that have done outstanding add the sphere of ML (AI) ar Facebook, Google, Microsoft, IBM, etc. that ar investment millions and billions during this terribly field of ML development and analysis. presently there's a large market and want for building Intelligent Systems for Recommendation. To counter this, one amongst the simplest and most desirable System is Recommendation System (RS). Recommendation Systems had well-tried to play a crucial role within the field of E-Commerce websites, on-line searching, chemical analysis Apps, Social-Networking, Digital selling, on- line Advertisements, etc. by providing customized recommends and feedback to users in step with their preferences and selections. the subject of this report is AI primarily based picture show Recommendation System. because the topic of this paper suggests we have a tendency to ar planning to discuss concerning varied ways in which and approaches of ML (AI) to make a picture show Recommendation System (RS) application. There ar several approaches to make a recommendation system in step with one’s want like cooperative Filtering, Content primarily based Recommendation Systems, K-Means algorithmic rule. which can be in brief mentioned during this report, however we are going to primarily discuss and work on K suggests that algorithmic rule approach

When Variety Seeking Meets Unexpectedness: Incorporating Variety-Seeking Behaviors into Design of Unexpected Recommender Systems

In this paper, we study the consumers’ variety-seeking behavior in recommender system applications and propose a comprehensive framework to measure such behavior based on past consumption records. The effectiveness of the proposed framework is validated through user questionnaire studies conducted at Alibaba, where our constructed variety-seeking measures match well with consumers’ self-reported levels of their variety-seeking behaviors. We subsequently present a recommendation framework that combines the identified variety-seeking levels with unexpected recommender systems in the data mining literature to address consumers’ heterogenous desire for product variety, in which we provide more unexpected product recommendations to variety-seeking consumers and vice versa. Through off-line experiments on three different recommendation scenarios and a large-scale online controlled experiment at a major video-streaming platform, we demonstrate that those models following our recommendation framework significantly increase various business performance metrics and generate tangible economic impact for the company. Our findings lead to important managerial implications to better understand consumers’ variety-seeking behaviors and design recommender systems. As a result, the best performing model in our proposed frameworks is deployed by the company to serve all consumers on the video-streaming platform.

An Optimal Algorithm for Finding Champions in Tournament Graphs

A tournament graph is a complete directed graph, which can be used to model a round-robin tournament between $n$ players. In this paper, we address the problem of finding a champion of the tournament, also known as Copeland winner, which is a player that wins the highest number of matches. In detail, we aim to investigate algorithms that find the champion by playing a low number of matches. Solving this problem allows us to speed up several Information Retrieval and Recommender System applications, including question answering, conversational search, etc. Indeed, these applications often search for the champion inducing a round-robin tournament among the players by employing a machine learning model to estimate who wins each pairwise comparison. Our contribution, thus, allows finding the champion by performing a low number of model inferences. We prove that any deterministic or randomized algorithm finding a champion with constant success probability requires $\Omega(\ell n)$ comparisons, where $\ell$ is the number of matches lost by the champion. We then present an asymptotically-optimal deterministic algorithm matching this lower bound without knowing $\ell$, and we extend our analysis to three variants of the problem. Lastly, we conduct a comprehensive experimental assessment of the proposed algorithms on a question answering task on public data. Results show that our proposed algorithms speed up the retrieval of the champion up to $13\times$ with respect to the state-of-the-art algorithm that perform the full tournament.

스마트 관광 활성화를 위한 트립어드바이저 애플리케이션 리뷰 분석

The development of information and communication technology and the improvement of the development and dissemination of smart devices have caused changes in the form of tourism, and the concept of smart tourism has since emerged. In this regard, researches related to smart tourism has been conducted in various fields such as policy implementation and surveys, but there is a lack of research on application reviews. This study collects Trip Advisor application review data in the Google Play Store to identify usage of the application and user satisfaction through Latent Dirichlet Allocation (LDA) topic modeling. The analysis results in four topics, two of which are positive and the other two are negative. We found that users were satisfied with the application's recommendation system, but were dissatisfied when the filters they set during search were not applied or that reviews were not published after updates of the application. We suggest more categories can be added to the application to provide users with different experiences. In addition, it is expected that user satisfaction can be improved by identifying problems within the application, including the filter function, and checking the application environment and resolving the error occurring during the application usage.

Comparative Analysis of Deep Learning Architectures and Vision Transformers for Musical Key Estimation

The musical key serves as a crucial element in a piece, offering vital insights into the tonal center, harmonic structure, and chord progressions while enabling tasks such as transposition and arrangement. Moreover, accurate key estimation finds practical applications in music recommendation systems and automatic music transcription, making it relevant across academic and industrial domains. This paper presents a comprehensive comparison between standard deep learning architectures and emerging vision transformers, leveraging their success in various domains. We evaluate their performance on a specific subset of the GTZAN dataset, analyzing six different deep learning models. Our results demonstrate that DenseNet, a conventional deep learning architecture, achieves remarkable accuracy of 91.64%, outperforming vision transformers. However, we delve deeper into the analysis to shed light on the temporal characteristics of each deep learning model. Notably, the vision transformer and SWIN transformer exhibit a slight decrease in overall performance (1.82% and 2.29%, respectively), yet they demonstrate superior performance in temporal metrics compared to the DenseNet architecture. The significance of our findings lies in their contribution to the field of musical key estimation, where accurate and efficient algorithms play a pivotal role. By examining the strengths and weaknesses of deep learning architectures and vision transformers, we can gain valuable insights for practical implementations, particularly in music recommendation systems and automatic music transcription. Our research provides a foundation for future advancements and encourages further exploration in this area.

Collaborative Filtering-Based Recommendation Systems for Touristic Businesses, Attractions, and Destinations

The success of touristic businesses, attractions, and destinations heavily relies on travel agents’ recommendations, which significantly impact client satisfaction. However, the underlying recommendation process employed by travel agents remains poorly understood. This study presents a conceptual model of the recommendation process and empirically investigates the influence of tourism categories on agents’ destination recommendations. By employing collaborative filtering-based recommendation systems and comparing various algorithms, including matrix factorization and deep learning models, such as the bilateral variational autoencoder (BiVAE) and light graph convolutional neural network, this research provides insights into the performance of different techniques in the context of tourism. The models were evaluated using a tourism dataset and assessed through a range of metrics. The results indicate that the BiVAE algorithm outperformed others in terms of ranking and prediction metrics, underscoring the significance of considering multiple measurements and exploring diverse techniques. The findings have practical implications for tourism marketers seeking to influence travel agents and offer valuable insights for researchers investigating this domain. Additionally, the proposed model holds potential for applications in travel recommendation systems, including attraction recommendations.

Elastic deep autoencoder for text embedding clustering by an improved graph regularization

Text clustering is a task for grouping extracted information of the text in different clusters, which has many applications in recommender systems, sentiment analysis, and more. Deep learning-based methods have become increasingly popular due to their high accuracy in identifying nonlinear structures. They usually consist of two major parts: dimensionality reduction and clustering. Autoencoders are simple unsupervised neural networks used for better representation of low-dimensional data and have shown good performance in dealing with non-linear features. However, while they utilize the Frobenius norm to deal well with Gaussian noise, they are sensitive to outlier data and Laplacian noise. In this paper, a deep autoencoder with an adapted elastic loss for text embedding clustering (EDA-TEC) is proposed. The elastic loss is a combination of the Frobenius norm and L2,1-norm to consider both types of noises. Additionally, to maintain the high-dimensional data geometric structure, a modified graph regularization term based on the weighted cosine similarity measure is used. EDA-TEC also improves clustering results by considering the sparsity regularization of the manifold representation data. In this jointly end-to-end deep learning model, better representation and text clustering results are achieved with high accuracy on common datasets compared to existing methods.11https://github.com/safinal/text-embedding-clustering.