Towards Faster Deep Collaborative Filtering via Hierarchical Decision Networks

Collaborative filtering (CF) is a class of recommendation methods that have been used in web contents such as video sites and electronic-commerce sites. To our knowledge, CF, however, has not been used in three-dimensional virtual worlds yet. In this paper, we verify the applicability of two state-of-the-art CF methods in a three-dimensional virtual world called Second Life (SL). Our experiment results on data from SL verify that CF methods can also be applied to area recommendation in virtual worlds. A discussion on which CF method should be used in this application domain is also given.

Collaborative filtering (CF) approach is successfully applied in the rating prediction of personal recommendation. But individual information source is leveraged in many of them, i.e., the information derived from single perspective is used in the user-item matrix for recommendation, such as user-based CF method mainly utilizing the information of user view, item-based CF method mainly exploiting the information of item view. In this paper, in order to take full advantage of multiple information sources embedded in user-item rating matrix, we proposed a rating-based integrated recommendation framework of CF approaches to improve the rating prediction accuracy. Firstly, as for the sparsity of the conventional item-based CF method, we improved it by fusing the inner similarity and outer similarity based on the local sparsity factor. Meanwhile, we also proposed the improved user-based CF method in line with the user-item-interest model (UIIM) by preliminary rating. Second, we put forward a background method called user-item-based improved CF (UIBCF-I), which utilizes the information source of both similar items and similar users, to smooth itembased and user-based CF methods. Lastly, we leveraged the three information sources and fused their corresponding ratings into an Integrated CF model (INTE-CF). Experiments demonstrate that the proposed rating-based INTE-CF indeed improves the prediction accuracy and has strong robustness and low sensitivity to sparsity of dataset by comparisons to other mainstream CF approaches.

In recent years, Recommender System (RS) research work has covered a wide variety of Artificial Intelligence techniques, ranging from traditional Matrix Factorization (MF) to complex Deep Neural Networks (DNN). Traditional Collaborative Filtering (CF) recommendation methods such as MF, have limited learning capabilities as it only considers the linear combination between user and item vectors. For learning non-linear relationships, methods like Neural Collaborative Filtering (NCF) incorporate DNN into CF methods. Though, CF methods still suffer from cold start and data sparsity. This paper proposes an improved hybrid-based RS, namely Neural Matrix Factorization++ (NeuMF++), for effectively learning user and item features to improve recommendation accuracy and alleviate cold start and data sparsity. NeuMF++ is proposed by incorporating effective latent representation into NeuMF via Stacked Denoising Autoencoders (SDAE). NeuMF++ can also be seen as the fusion of GMF++ and MLP++. NeuMF is an NCF framework which associates with GMF (Generalized Matrix Factorization) and MLP (Multilayer Perceptrons). NeuMF achieves state-of-the-art results due to the integration of GMF linearity and MLP non-linearity. Concurrently, incorporating latent representations has shown tremendous improvement in GMF and MLP, which result in GMF++ and MLP++. Latent representation obtained through the SDAEs' latent space allows NeuMF++ to effectively learn user and item features, significantly enhancing its learning capability. However, sharing feature extractions among GMF++ and MLP++ in NeuMF++ might hinder its performance. Hence, allowing GMF++ and MLP++ to learn separate features provides more flexibility and greatly improves its performance. Experiments performed on a real-world dataset have demonstrated that NeuMF++ achieves an outstanding result of a test root-mean-square error of 0.8681. In future work, we can extend NeuMF++ by introducing other auxiliary information like text or images. Different neural network building blocks can also be integrated into NeuMF++ to form a more robust recommendation model.

Cooking your own food at home is a good step toward reducing fast food consumption. Fast food increases the risk of dangerous diseases. The diversity of recipe information available on the internet makes it difficult to choose recipes that match user preferences. Mobile technology can help with this by recommending recipes that better suit users' eating habits. This makes the transition to a healthier diet easier. Therefore, in this study, a recommendation system was developed that can recommend recipes based on the preferences of Android users. Two main recommendation methods are used in this study: content-based filtering and collaborative filtering. Using cosine similarity, a content-based recommendation system identifies the proximity between a recipe for food and its related context. The history of user comments on recipes serves as implicit feedback for the collaborative recommendation algorithm. This eliminates the need for explicit evaluations, such as ratings. This recommendation system generates recommendations in the form of the top ten food recipes with an evaluation matrix, referred to as NDCG@k and Hit-Ratio@k. The tests revealed that a content-based filtering technique may produce helpful recommendations, with the highest similarity score of 0.41 for the entry "chocolate cake that you can easily make at home." Meanwhile, in the collaborative filtering method using the Neural Collaborative Filtering (NCF) approach, the system shows consistent performance improvements, with the MAP@10 value increasing from 0.705 to 0.767 and the NDCG@10 from 0.78 to 0.83 after 10 training epochs. Keywords: Recommendation systems; content-based filtering; neural collaborative filtering; cosine similarity; implicit feedback

In this paper, a new two levels propagated neighborhoods based collaborative filtering method (PNCF) is proposed for developing effective and efficient recommendation system. Traditional collaborative filtering (CF) algorithms focus on construct k-nearest neighborhood for each item/user from user-item purchase/rating matrix, such as item-based k-nearest-neighbor collaborative filtering method (itemKNN) and user-based k-nearest-neighbor collaborative filtering method (userKNN). However, the utilization of K-nearest neighborhood method for singe item/user always misses some nature neighbors due to inevitable data noise and data sparsity, resulting in poor prediction accuracy. A novel two levels propagated neighborhoods construction strategy is introduced in PNCF to complement traditional K-nearest neighborhood method, uncovering the underlying neighborhood relationship of each data sample. Furthermore, utilizing propagated neighborhoods improves the recommendation quality. Numerous experiments on MovieLens data set show the superiority of our approach over current state of the art recommendation methods.

Utilizing transfer learning for in-domain collaborative filtering

The research of link prediction is widely used in many fields such as social relations and biological sciences. So it has become a very important research direction. Existing link prediction methods are usually based on the assumption that "the more similar two nodes are, the higher the probability that a link exists between them". That is, the similarity between nodes is considered as the existence probability of a link between nodes. Such an approach does not fully exploit the hidden information in the network, so the accuracy of prediction needs to be further improved. In this paper, we propose a link prediction model based on collaborative filtering. The model combines the similarities between nodes and the known network topology by the method of collaborative filtering, which effectively improves the accuracy of link prediction. Extensive experiments on six real-world datasets demonstrate the validity of the proposed model.

Towards similar alignment and unique uniformity in collaborative filtering

Internet sources contain a vast amount of information about items that people desire to purchase. It is impossible to evaluate these resources and come to an informed decision. People need automated systems that evaluate previous information and propose item alternatives. Recommending items using a smart system, which is based on the previous user preferences, has growing importance since the available product data is exponentially growing. Additionally, it is difficult to find new and correct things that a user would like among this massive amount of data. To make accurate recommendations with a smart system, researchers and practitioners use collaborative filtering methods with similarity calculation based on user preferences. The crucial point in collaborative filtering is to find a valuable measure that resembles correct similarity between users. The current similarity metrics in the literature have some disadvantages in conducting accurate recommendations. To improve the recommendation performance, this study proposes a novel similarity measure that assesses the distance between the user’s ratings and the median score. Considering distance from the median score is essential since some users may prefer to rate close to the median rather than the extremes. Experiments were conducted with a famous collaborative filtering dataset. Results showed that proposed similarity measure demonstrated superior performance regarding the recommendation accuracy. Implications of our results for XYZ are discussed.

This study proposes a hybrid recommendation framework for online shopping platforms that integrates Convolutional Neural Networks (CNNs) with collaborative filtering (CF) to improve personalization accuracy. Traditional recommendation approaches, such as content-based and collaborative filtering methods, often suffer from data sparsity, cold-start problems, and limited capability to capture rich item features, which reduces recommendation effectiveness in large-scale e-commerce systems. To overcome these challenges, the proposed model leverages CNNs to extract visual and textual features of products while CF learns latent user–item interaction patterns. By combining content-level representations with hidden user preferences and engagement behavior, the hybrid approach enhances recommendation relevance and accuracy. The model was trained and evaluated using the Kaggle E-Commerce Product Recommendation Collaborative dataset. Experimental results demonstrate superior prediction performance compared to conventional methods, highlighting improved scalability, accuracy, and user engagement. The findings confirm the effectiveness of the CNN-based collaborative filtering framework as a robust and data-driven solution for modern e-commerce recommendation systems, enabling more personalized and relevant shopping experiences.

In collaborative filtering, many neighbors are needed to improve the quality and stability of the recommendation. The quality may not be good mainly due to the high similarity between two users not guaranteeing the same preference for products considered for recommendation. This paper proposes a consistency definition, rather than similarity, based on information entropy between two users to improve the recommendation. This kind of consistency between two users is then employed as a trust metric in collaborative filtering methods that select neighbors based on the metric. Empirical studies show that such collaborative filtering reduces the number of neighbors required to make the recommendation quality stable. Recommendation quality is also significantly improved.

Online job boards are one of the central components of modern recruitment industry. With millions of candidates browsing through job postings everyday, the need for accurate, effective, meaningful, and transparent job recommendations is apparent more than ever. While recommendation systems are successfully advancing in variety of online domains by creating social and commercial value, the job recommendation domain is less explored. Existing systems are mostly focused on content analysis of resumes and job descriptions, relying heavily on the accuracy and coverage of the semantic analysis and modeling of the content in which case, they end up usually suffering from rigidity and the lack of implicit semantic relations that are uncovered from users' behavior and could be captured by Collaborative Filtering (CF) methods. Few works which utilize CF do not address the scalability challenges of real-world systems and the problem of cold-start. In this paper, we propose a scalable item-based recommendation system for online job recommendations. Our approach overcomes the major challenges of sparsity and scalability by leveraging a directed graph of jobs connected by multi-edges representing various behavioral and contextual similarity signals. The short lived nature of the items (jobs) in the system and the rapid rate in which new users and jobs enter the system make the cold-start a serious problem hindering CF methods. We address this problem by harnessing the power of deep learning in addition to user behavior to serve hybrid recommendations. Our technique has been leveraged by CareerBuilder.com which is one of the largest job boards in the world to generate high-quality recommendations for millions of users.

Matrix factorization in Cross-domain Recommendations Framework by Shared Users Latent Factors

In order to tackle the problem of information overload and effective recommendation based on users' preference, need, and interest a number of research contributions has been made for the development of recommender systems. However, certain challenges, such as data sparsity, profiling attack, and black-box recommendation still exist and hamper their prediction accuracy. In this paper, we propose a hybrid approach to predict user ratings by incorporating both trust and context of the users in traditional recommender systems using collaborative filtering method. The similarity between two users is computed using both trust value and context-based similarity. The trust value is based on three trust statements -- rating deviation, emotions, and reviews helpfulness. On the other hand, context-based similarity is based on four contextual features -- companion, place, day, and priority. The performance of the proposed trust- and context-based hybrid approach is analyzed using mean absolute error and root mean square error on a real dataset generated from two movie data sources (IMDB and Rotten Tomatoes), and it performs significantly better in comparison to some of the standard baseline methods. The rating prediction using only trust statements gives better results in comparison to other collaborative filtering approaches, such as user-based and item-based filtering approaches. Similarly, context-based collaborative filtering approach also outperforms standard collaborative filtering approaches. In addition, rating prediction using both trust- and context-based features performs better in comparison to only trust-based or context-based approaches.

Collaborative filtering is the most popular approach to build recommender systems and has been successfully employed in many applications. However, it cannot make recommendations for so-called cold start users that have rated only a very small number of items. In addition, these methods do not know how confident they are in their recommendations. Trust-based recommendation methods assume the additional knowledge of a trust network among users and can better deal with cold start users, since users only need to be simply connected to the trust network. On the other hand, the sparsity of the user item ratings forces the trust-based approach to consider ratings of indirect neighbors that are only weakly trusted, which may decrease its precision. In order to find a good trade-off, we propose a random walk model combining the trust-based and the collaborative filtering approach for recommendation. The random walk model allows us to define and to measure the confidence of a recommendation. We performed an evaluation on the Epinions dataset and compared our model with existing trust-based and collaborative filtering methods.