A Knowledge Graph-Enhanced Attention Aggregation Network for Making Recommendations

In recent years, many researchers have devoted time to designing algorithms used to introduce external information from knowledge graphs, to solve the problems of data sparseness and the cold start, and thus improve the performance of recommendation systems. Inspired by these studies, we proposed KANR, a knowledge graph-enhanced attention aggregation network for making recommendations. This is an end-to-end deep learning model using knowledge graph embedding to enhance the attention aggregation network for making recommendations. It consists of three main parts. The first is the attention aggregation network, which collect the user’s interaction history and captures the user’s preference for each item. The second is the knowledge graph-embedded model, which aims to integrate the knowledge. The semantic information of the nodes and edges in the graph is mapped to the low-dimensional vector space. The final part is the information interaction unit, which is used for fusing the features of two vectors. Experiments showed that our model achieved a stable improvement compared to the baseline model in making recommendations for movies, books, and music.

An Intelligent Question Answering Platform for Graduate Enrollment

To enhance the competitiveness of colleges and universities in the graduate enrollment and reduce the pressure on candidates for examination and consultation, it is necessary and practically significant to develop an intelligent Q&A platform, which can understand and analyze users' semantics and accurately return the information they need. However, there are problems such as the low volume and low quality of the corpus in the graduate enrollment, this paper develops a question answering platform based on a novel retrieval model including density-based logistic regression and the combination of convolutional neural networks and bidirectional long short-term memory. The experimental results show that the proposed model can effectively alleviate the problem of data sparseness and greatly improve the accuracy of the retrieval performance for the graduate enrollment.

Recommendations of Insurance Products based on User Behaviour

The remarkable increase in competition within the insurance sector has resulted in an overwhelming number of insurance products being available in the market. With rapid development of recommendation system, how to accurately predict Insurance policies using user lifestyle choices has become more and more important. The problem with the traditional systems is data sparseness. This paper proposes a recommender system to predict insurance products for new and existing customers. The main goal of the proposed system is to generate personalized recommendations based on the user lifestyle practices. By providing accurate personalised recommendations, the customer experience with the insurers can be improved.

Deep Personalized Medical Recommendations Based on the Integration of Rating Features and Review Sentiment Analysis

To comply with the rapid development of big data in mobile services, an increasing number of websites have begun to provide users with recommendation decisions in various areas, like shopping, tourism, food, and medical treatment. However, there are still some challenges in the field of medical recommendation systems, such as the lack of personalized medical recommendations and the problem of data sparseness, which seriously restricts the effectiveness of such recommendations. In this paper, we propose a personalized medical recommendation method based on a convolutional neural network that integrates revised ratings and review text, called revised rating and review based on a convolutional neural network (RR&R-CNN). First, the review text is divided into user and doctor datasets, and BERT vectorized representations are performed on them. Moreover, the original rating features are revised by adding the sentiment analysis values of the review text. Then, the vectorized review text and the revised rating features are spliced together and input into the convolutional neural network to extract the deep nonlinear feature vectors of both users and doctors. Finally, we use a factorization machine for feature interaction. We conduct comparison experiments based on a Yelp dataset in the “Health & Medical” category. The experimental results confirm the conclusion that RR&R-CNN has a better effect compared to a traditional method.

Uncovering marine connectivity through sea surface temperature

A foundational paradigm in marine ecology is that Oceans are divided into distinct ecoregions demarking unique assemblages of species where the characteristics of water masses, and quantity and quality of environmental resources are generally similar. In most of the world Ocean, defining these ecoregions is complicated by data sparseness away of coastal areas and by the large-scale dispersal potential of ocean currents. Furthermore, ocean currents and water characteristics change in space and time on scales pertinent to the transitions of biological communities, and predictions of community susceptibility to these changes remain elusive. Given recent advances in data availability from satellite observations that are indirectly related to ocean currents, we are now poised to define ecoregions that meaningfully delimit marine biological communities based on their connectivity and to follow their evolution over time. Through a time-dependent complex network framework applied to a thirty-year long dataset of sea surface temperatures over the Mediterranean Sea, we provide compelling evidence that ocean ecoregionalization based on connectivity can be achieved at spatial and time scales relevant to conservation management and planning.

A review of recommendation system research based on bipartite graph

The interaction history between users and items is usually stored and displayed in the form of bipartite graphs. Neural network recommendation based on the user-item bipartite graph has a significant effect on alleviating the long-standing data sparseness and cold start of the recommendation system. The whole paper is based on the bipartite graph. An review of the recommendation system of graphs summarizes the three characteristics of graph neural network processing bipartite graph data in the recommendation field: interchangeability, Multi-hop transportability, and strong interpretability. The biggest contribution of the full paper is that it summarizes the general framework of graph neural network processing bipartite graph recommendation from the models with the best recommendation effect in the past three years: embedding layer, propagation update layer, and prediction layer. Although there are subtle differences between different models, they are all this framework can be applied, and different models can be regarded as variants of this general model, that is, other models are fine-tuned on the basis of this framework. At the end of the paper, the latest research progress is introduced, and the main challenges and research priorities that will be faced in the future are pointed out.

3D reconstruction of genomic regions from sparse interaction data

Chromosome Conformation Capture (3C) technologies measure the interaction frequency between pairs of chromatin regions within the nucleus in a cell or a population of cells. Some of these 3C technologies retrieve interactions involving non-contiguous sets of loci, resulting in sparse interaction matrices. One of such 3C technologies is Promoter Capture Hi-C (pcHi-C) that is tailored to probe only interactions involving gene promoters. As such, pcHi-C provides sparse interaction matrices that are suitable to characterise short- and long-range enhancer-promoter interactions. Here, we introduce a new method to reconstruct the chromatin structural (3D) organisation from sparse 3C-based datasets such as pcHi-C. Our method allows for data normalisation, detection of significant interactions, and reconstruction of the full 3D organisation of the genomic region despite of the data sparseness. Specifically, it produces reliable reconstructions, in line with the ones obtained from dense interaction matrices, with as low as the 2-3% of the data from the matrix. Furthermore, the method is sensitive enough to detect cell-type-specific 3D organisational features such as the formation of different networks of active gene communities.

Prior information for population pharmacokinetic and pharmacokinetic/pharmacodynamic analysis: overview and guidance with a focus on the NONMEM PRIOR subroutine

Population pharmacokinetic analysis is used to estimate pharmacokinetic parameters and their variability from concentration data. Due to data sparseness issues, available datasets often do not allow the estimation of all parameters of the suitable model. The PRIOR subroutine in NONMEM supports the estimation of some or all parameters with values from previous models, as an alternative to fixing them or adding data to the dataset. From a literature review, the best practices were compiled to provide a practical guidance for the use of the PRIOR subroutine in NONMEM. Thirty-three articles reported the use of the PRIOR subroutine in NONMEM, mostly in special populations. This approach allowed fast, stable and satisfying modelling. The guidance provides general advice on how to select the most appropriate reference model when there are several previous models available, and to implement and weight the selected parameter values in the PRIOR function. On the model built with PRIOR, the similarity of estimates with the ones of the reference model and the sensitivity of the model to the PRIOR values should be checked. Covariates could be implemented a priori (from the reference model) or a posteriori, only on parameters estimated without prior (search for new covariates). Graphic abstract

RESEARCH ON STUDENT BEHAVIOR INFERENCE METHOD BASED ON FP-GROWTH ALGORITHM

How to use modern information technology to efficiently and quickly obtain the personalized recommendation information required by students, and to provide high-quality intelligent services for schools, parents and students has become one of the hot issues in college research. This paper uses FP-growth association rule mining algorithm to infer student behavior and then use the collaborative filtering recommendation method to push information according to the inference result, and then push real-time and effective personalized information for students. The experimental results show that an improved FP-growth algorithm is proposed based on the classification of students. The algorithm combines the student behavior inference method of FP-growth algorithm with the collaborative filtering hybrid recommendation method, which not only solves the FP-tree tree branch. Excessive and collaborative filtering recommendation algorithm data sparseness problem, can also analyze different students' behaviors and activities, and accurately push real-time, accurate and effective personalized information for students, to promote smart campus and information intelligence The development provides better service.

Punctuation and Parallel Corpus Based Word Embedding Model for Low-Resource Languages

To overcome the data sparseness in word embedding trained in low-resource languages, we propose a punctuation and parallel corpus based word embedding model. In particular, we generate the global word-pair co-occurrence matrix with the punctuation-based distance attenuation function, and integrate it with the intermediate word vectors generated from the small-scale bilingual parallel corpus to train word embedding. Experimental results show that compared with several widely used baseline models such as GloVe and Word2vec, our model improves the performance of word embedding for low-resource language significantly. Trained on the restricted-scale English-Chinese corpus, our model has improved by 0.71 percentage points in the word analogy task, and achieved the best results in all of the word similarity tasks.