Search

To address the problem of insufficient recommendation performance in session recommendation, this paper proposed a disentangled graph neural network model (GDST-GNN) using global graph diffusion and spatio-temporal awareness. Specifically, the model firstly constructed a global collaborative graph based on all sessions from a global perspective, and then employed graph diffusion as the message propagation paradigm for global representation learning of items to capture global information be- yond the current session. For the representation learning of the current session, this paper designed a disentangled spatio-temporal gated network modeling the complex transition pattern and temporal dependency pattern of items in the session, and then fused the learned global and local representations factor by factor. In addition, this paper employed a self-supervised task to achieve performance enhancement of the model. Finally, it generated session representations through attention networks to achieve accurate recommendation of items. Extensive experiments on four real-world datasets validate the effectiveness of the proposed model. [ABSTRACT FROM AUTHOR]

Various aquatic diseases occur frequently in recent years, particularly with the rapid development of the aquaculture industry. Once these diseases spread very rapidly, a very serious risk can be put on aquaculture. Furthermore, the network data of aquatic diseases also presents highly dispersed, multi-source heterogeneous features with the development of Internet technology. It is a high demand to rapidly and accurately obtain the required information, due to the explosive growth of network data. However, traditional information acquisition cannot fully meet the search engines. The retrieval keywords or shallow semantic analysis can bring a large number of related web links, leading to vague and redundant answers. The intelligent Q&A system can be selected to support the users' natural language input, and then accurately capture the user intent, finally returning concise and accurate answers. Among them, the emergence and rapid development of knowledge graphs can provide a high-quality knowledge base for intelligent question-answering systems, in order to promote the application of question-answering systems in various fields. The knowledge graph construction can be divided mainly into four steps: data acquisition, ontology construction, knowledge extraction and storage. Firstly, the crawler technology is used to obtain the relevant aquatic disease data, and then data preprocessing can be performed, including data cleaning and analysis. Secondly, the aquatic diseases ontology can be constructed using the data content and representation characteristics, in order to predefine the relations and properties types between entities. As such, the boundaries of knowledge extraction are clarified during this time. Secondly, the rule logic can be used to extract the semi-structured data. The entity and relation joint extraction is then used to extract unstructured data. Finally, the extracted triple data is stored in the Neo4j graph database, in order to realize the visual management of the knowledge graph and a certain degree of knowledge reasoning. In this study, a new text annotation system of "aquatic disease + relationship +BMES" was proposed in the unstructured knowledge. The relationship extraction was also integrated into the named entity recognition task. Then, the ternary model was directly constructed to transform the entity relationship extraction into sequence annotation, in order to improve the annotation efficiency at least twice for the joint extraction of entity and relationship. At the same time, the triplet data was obtained for the triplet-building module using label matching and mapping. The overlapping relation extraction was then solved in this case. The BERT-BiLSTM+CRF end-to-end model was used to carry out the test. The test results showed that the triad extraction shared a high recall rate (89.64%), accuracy (94.04%), and F1 (91.34%), which was significantly better than CNN+BiLSTM+CRF and BiLSTM+CRF models. The extracted knowledge was stored in the Neo4j graph database, and then realized knowledge visualization management and knowledge reasoning analysis. The aquatic disease knowledge map presented high precision and fine granularity. The finding can provide a new idea for the field of intelligent Q&A. Anyway, the semi-automatic construction of a knowledge graph can also offer technical support for the recommendation system, knowledge base construction, search and application knowledge base construction. [ABSTRACT FROM AUTHOR]

Recommendation system is one of the most common and basic applications in E-Commerce, search engines, etc. However, the user-item rating matrix data in the recommendation system is extremely sparse, resulting in poor recommendation results. This paper proposes a dual auto encoders matrix factorization (DAE-MF) model for the sparseness of the scoring matrix. This model uses a two-way parallel network structure, which combines a convolutional auto encoder and stacked denoising autoencoder and matrix factorization technology, fully combining the advantages of the two modules, effectively improving the model’s data completion capabilities. Through experiments on MovieLens and AIV datasets, it is verified that the DAE-MF method is superior to the existing models in rating prediction tasks. [ABSTRACT FROM AUTHOR]

Expert recommendation is one of the research hotspots in the online Q&A community. However, the majority of the currently used algorithms concentrate on matching user's static interest and question information, ignoring the efficient capture of user' s dynamic interest representation information, which results in a lack of recommendation accuracy. In order to address the aforementioned issues, this paper proposed an expert recommendation algorithm for the Q&A community that incorporated multi-head self-attention. Firstly, this algorithm built out question encoders of convolutional neural networks and attention mechanisms to process target question and user' s historically answered questions, in order to extract the corresponding question representations. Secondly, it treated the sequence of user's historically answered questions as a time series and used the multi-head self-attention mechanism to learn the dynamic interest representations hidden in the sequence, and combined with the user' s static interest representations to obtain the user' s comprehensive interest representations. Finally, it calculated the similarity between the target question representation and the comprehensive user representation to generate recommendation results. Based on the real data from the Zhihu Q&A community, the experimental results show that the performance of this algorithm is significantly better than the current popular expert recommendation algorithm. [ABSTRACT FROM AUTHOR]

Traditional recommendation algorithms map text content through topic model or mean value of word vectorization. For the issue that existing methods cannot make full use of text information or ignore word order information, this study proposes a multi-task learning recommendation method for scientific literature. Based on the multi-task learning framework, an encoder is designed and a GL model is established. The GL model is trained to combine content recommendation and text metadata prediction, which improves the sparsity of traditional collaborative filtering and regularizes the collaborative filtering model. Finally, an evaluation test is carried out on public and private data sets respectively, and the superiority of the proposed method is demonstrated by comparing with the existing classical methods. [ABSTRACT FROM AUTHOR]

Recommender system can find information that satisfies the user′s individual needs from huge amount of information. With the development of deep learning, deep learning has been widely applied in recommender systems. CTR prediction plays an important role in recommender system and has been widely used in many fields such as personalized recommendation, information retrieval, online advertising and so on. For the issue of large and sparse data in recommender system, this study merges xDeepFM model with attention network, and proposes a new CTR prediction model based on deep learning, which is called Atte-xDeepFM model. This model can solve the issue of feature scarcity, effectively learn the interactions relationship between features, and does not need to manually extract useful information in feature engineering. The comparative experiments on Avazu and Criteo data sets prove the effectiveness of the proposed model. Compared with the algorithm model commonly used in CTR prediction, the proposed model has better recommendation effect. [ABSTRACT FROM AUTHOR]

To address the problem of poor recommendation accuracy in the session recommendation model, this paper proposed a capsule-based graph convolutional disentanglement for session-aware recommendation (CGCD) method.Specifically, it used disentangled learning technique to transition item embedding into factor embedding based on multiple sub-channels. Then, it used a capsule dynamic fusion strategy to aggregate different factors to obtain a new item embedding.In addition, it used a multi-head attention mechanism to assign weights to each item in the session. Finally, it aggregated the item embedding with other items in the current session according to the assigned weights, and then generated an accurate session representation to achieve item recommendation. Experiments on two publicly available real datasets show that the proposed model can improve the Pre@10, Pre@20, MRR@10 and MRR@20 by 5.17%, 2.99%, 6.56% and 2.94% on average of the recommended performance, which verifies the effectiveness and efficiency of the method in this paper. [ABSTRACT FROM AUTHOR]

In order to improve the safety of pedestrian interaction in complex traffic scenes, this paper proposes a pedestrian trajectory prediction algorithm SAN-GAN (Social Angle Norm-GAN) based on Social-GAN network (Social-Generative Adversarial Network). The algorithm first takes the pedestrian’s historical location information and head information as input, obtains the hidden feature information of the pedestrian through the trajectory generator LSTM network (Long Short Term Memory Networks), captures the dynamic changes of the pedestrian’s visual field based on the pedestrian visual field module, builds a fan-shaped visual field for all pedestrians and filters the valid information, thus driving the neural network model to predict future trajectory changes of pedestrians. By comparing SAN-GAN with LSTM, Social-LSTM (Social-Long Short Term Memory Networks), Social-GAN and other trajectory prediction algorithms, The results show that the SAN-GAN algorithm reduces the ADE by an average of 65.8%, 51.2% and 10.7%, and the FDE by an average of 73.6%, 60.9% and 10.4%, respectively, in predicting the pedestrian trajectory for 3.2 s. The SAN-GAN is effective in predicting the future trajectory of pedestrians interacting in complex traffic environments compared to other algorithms. [ABSTRACT FROM AUTHOR]

With the advances in artificial intelligence, machines can now integrate information and extract its features to build a predictive model of likely outcomes, providing more insights to make better decisions. An artificial neural network, which is based on the structure and functions of the human brain, is a series of algorithms that can filter information into layers of recognizable characteristics. The computational power of neural networks is increasing in parallel with improved development of hardware devices. Artificial intelligence has become inseparable from our daily lives because it assists us in many areas. In recent years, researchers in sports have begun to incorporate the tools of artificial intelligence to complete many tasks that used to rely on substantial labor inputs. For instance, player and object trajectory tracking makes it easier to understand team play tactics and to analyze the strategies used in games. Motion recognition provides broadcasting and live-streaming programs with quick information on what skills are deployed. Data analysis can prevent athletes from sustaining injuries. Thanks to the continual improvements in artificial intelligence, more and more studies on its application in sports are expected. Language-teaching and referee robots are likely to be introduced in the near future. [ABSTRACT FROM AUTHOR]

The recommendation algorithm based on graph neural network can extract the association relationship between users and goods. Traditional methods can’t extract this relationship. At present, most of these algorithms ignore the general prefe-rences in the review data of users and products. In order to solve this problem, this paper proposed a new method. This me-thod used the graph neural network to extract association relations, and the advantage of deep learning to extract the general preferences, and carried out feature fusion to improve the recommendation effect. This paper conducted comparative experiments and ablation experiments on four sets of public data sets to verify the effectiveness of the proposed method. The evaluation indexes include the recall rate and normalized discounted cumulative gain. Experiments show that this method is more effective than the existing algorithms. The feature fusion of the two networks can improve the recommendation effect. [ABSTRACT FROM AUTHOR]

In recent years, as deep learning has achieved good results in many fields, deep learning has also begun to be applied to recommendation systems, such as NFM models and DeepFM models that use deep learning technology to capture high-level feature interactions. However, considering the changes in the external environment and internal perception, the user' s interest should also change dynamically over time, and the combination based on the original features may not necessarily learn effective feature interaction. This paper attempted to build a new model FG_DRFwFm, which could learn the interaction of low-level and high-level features of multiple feature domains and dealt with long-term changes in user interest. It constructed the training features by constructing new features based on the original features and splicing them together, which could better learn effective feature interaction. Finally, the proposed model compared the recommendation effect with multiple advanced CTR algorithms on the MovieLens data set. The experimental results show that the proposed model achieves better results. [ABSTRACT FROM AUTHOR]

Deep matrix factorization adopts deep nonlinear mapping to break through the bottleneck of bilinear relationship affecting the performance of the recommendation system in matrix factorization. However, it doesn't take into account users' preference for unrated items, and its recommendation performance isn't advantageous for large-scale data with high sparsity. This paper proposed a recommendation algorithm for fusing matrix completion and deep matrix factorization. First, it filled the unknown elements in the original scoring matrix by the matrix completion model. Then, based on the completed matrix, it constructed the latent vectors of users and items by deep learning model. Finally, this paper tested the proposed model on the MovieLens and SUSHI datasets. Experimental results show that the proposed algorithm significantly improves the performance of the recommendation system compared to deep matrix factorization. [ABSTRACT FROM AUTHOR]

At present, in the recommendation task based on document information, the traditional document-based hybrid recommendation algorithm still relies on a shallow linear model. When the scoring data becomes huge and complex, its recommendation performance is often not ideal. This paper proposed a deep fusion model( DeepFM), which could fully capture text information and handle complex and sparse scoring data well. DeepFM consisted of two parallel neural networks . One of them used multiple layers of perceptron to extract the row vector information of the rating matrix to obtain the user' s potential feature vector. The other one used fusion model of MLP and convolutional neural network (CNN) to extract additional textual information about the item and obtain the item' s potential feature vector. Finally, it fused the user feature vector and the item feature vector by constructing a fusion layer to obtain a prediction score. Experimental results show that DeepFM outperforms mainstream recommendation models on MovieLens dataset and Amazon dataset. [ABSTRACT FROM AUTHOR]

In the recommendation system. the inner product interaction of a single learning matrix de composition or the use of deep neural networks to capture user interaction with the project is not sufficient to effectively learn the potential characteristics of users and projects. In view of this problem. this paper proposes a collaborative filtering algorithm based on deep learning feature representation (DLFeaCF) based on display feedback and implicit feedback. The model first learns the inner product and outer product interaction between the user and the project. Then, based on the inner product, it uses the nonlinear interactive learning ability of the multi-layer perceptron (MLP) to obtain the two aspects from implicit mapping and feature mapping: on the basis of the outer product, CNN learning is used to capture the local features of the user and the project finally, combine the features in the fusion layer and obtain the prediction score, The experiments on the real MovieLens dataset show that the DLFeaCF model can achieve better recommendation performance. [ABSTRACT FROM AUTHOR]

In order to suppress the disturbance of auxiliary information in various directions in the recommendation model and consider using text information to extract item features，this paper proposes a matrix factorization model，which is a SVD++ model of hybrid deep collaborative filtering for recommender method (SHD). The model combines additional stacked denoising autoencoder (ASDAE) and stacked contractive denoising autoencoder (CDAE) with auxiliary information to extract latent feather representations of users and items，and adds pre-trained word embedding model to consider the semantic relationship of words when extracting item feature representations. Experimental results on the MovieLens_1M and MovieLens_10M show that compared with traditional algorithms, deep learning algorithms, and variants of the proposed model, the proposed model is more effective in extracting latent feature representations and improving the accuracy of prediction score. [ABSTRACT FROM AUTHOR]

Point-of-interest (POI) recommendation has emerged as a focal point in the research of location-based social network (LBSN) in recent years. It can help users find their favorite venue and bring considerable benefits to businesses. Nowadays, deep learning is gradually applied to the task of recommendation system because it can capture the nonlinear relationship between users and items more effectively. This paper thus focuses on recent research on POI recommendation combined with deep learning. Firstly, we introduce the difference between POI recommendation and other traditional recommendation tasks and illustrate various influencing factors that can improve the performance of the model. Then, the methods of applying deep learning to POI recommendation are divided into four categories, including POI embedding, deep collaborative filtering, feature extraction from side information, and sequence recommendation using recurrent neural network (RNN). We also investigate the development of user models performance and advantages combined with deep learning in these different aspects of applications. Finally, we summarize and look forward to the development of POI recommendation research combined with deep learning. [ABSTRACT FROM AUTHOR]

Recommender systems can effectively reduce the information overload by recommending items that users may be interested in. Euclidean-embedding-based collaborative filtering methods map users and items to a unified latent space, which is one of the most important methods to build a recommender system. However, traditional Euclidean-embedding-based collaborative filtering methods only consider the low-order interaction between user latent feature vectors and item latent feature vectors, and cannot efficiently model the complex interaction behavior between users and items in the real world. In this paper, we propose a deep-Euclidean-embedding-based collaborative filtering algorithm, which utilizes deep learning technology to learn the high-order and nonlinear interaction function between user latent feature vectors and item latent feature vectors. This can model the complex interaction behavior between users and items. Experimental results on real-world datasets show that our proposed algorithm outperforms traditional collaborative filtering algorithms. [ABSTRACT FROM AUTHOR]

Most of the existing modeling methods of combining metadata are based on the same user I item attribute weights, so that the key relationships between users and items are not significant, and it is difficult to obtain better recommendation performance. To solve the above problems, this paper proposed a method of deep joint learning recommendation based on metadata and attention mechanism. It used double deep network joint learning, one of the networks implemented matrix nonlinear decomposition based on implicit feedback data to learn user/ project personalization relationship, and another network automatically captured influence the user/ item key attributes to recommend by using attention mechanism, through the user preference relation with weighted of different attributes modeling highlighted the extended model. Experimental results show that the proposed recommendation algorithm has better recommendation performance on two public datasets of MovieLens 100K and MovieLens lM. [ABSTRACT FROM AUTHOR]