The subjects of literature are the direct expression of the author’s research results. Mining valuable knowledge helps to save time for the readers to understand the content and direction of the literature quickly. Therefore, the co-occurrence network of high-frequency words in the bioinformatics literature and its structural characteristics and evolution were analysed in this paper. First, 242,891 articles from 47 top bioinformatics periodicals were chosen as the object of the study. Second, the co-occurrence relationship among high-frequency words of these articles was analysed by word segmentation and high-frequency word selection. Then, a co-occurrence network of high-frequency words in bioinformatics literature was built. Finally, the conclusions were drawn by analysing its structural characteristics and evolution. The results showed that the co-occurrence network of high-frequency words in the bioinformatics literature was a small-world network with scale-free distribution, rich-club phenomenon and disassortative matching characteristics. At the same time, the high-frequency words used by authors changed little in 2–3 years but varied greatly in four years because of the influence of the state-of-the-art technology.

The recommendation algorithm in e-commerce systems is faced with the problem of high sparsity of users’ score data and interest’s shift, which greatly affects the performance of recommendation. Hence, a combined recommendation algorithm based on improved similarity and forgetting curve is proposed. Firstly, the Pearson similarity is improved by a wide range of weighted factors to enhance the quality of Pearson similarity for high sparse data. Secondly, the Ebbinghaus forgetting curve is introduced to track a user’s interest shift. User score is weighted according to the residual memory of forgetting function. Users’ interest changing with time is tracked by scoring, which increases both accuracy of recommendation algorithm and users’ satisfaction. The two algorithms are then combined together. Finally, the MovieLens dataset is employed to evaluate different algorithms and results show that the proposed algorithm decreases mean absolute error (MAE) by 12.2%, average coverage 1.41%, and increases average precision by 10.52%, respectively.

The unobserved heterogeneity in traffic crash data hides certain relationships between the contributory factors and injury severity. The literature has been limited in exploring different types of clustering methods for the analysis of the injury severity in crashes involving large trucks. Additionally, the variability of data type in traffic crash data has rarely been addressed. This study explored the application of the k-prototypes clustering method to countermeasure the unobserved heterogeneity in large truck-involved crashes that had occurred in the United States between the period of 2016 to 2019. The study segmented the entire dataset (EDS) into three homogeneous clusters. Four gradient boosted decision trees (GBDT) models were developed on the EDS and individual clusters to predict the injury severity in crashes involving large trucks. The list of input features included crash characteristics, truck characteristics, roadway attributes, time and location of the crash, and environmental factors. Each cluster-based GBDT model was compared with the EDS-based model. Two of the three cluster-based models showed significant improvement in their predicting performances. Additionally, feature analysis using the SHAP (Shapley additive explanations) method identified few new important features in each cluster and showed that some features have a different degree of effects on severe injuries in the individual clusters. The current study concluded that the k-prototypes clustering-based GBDT model is a promising approach to reveal hidden insights, which can be used to improve safety measures, roadway conditions and policies for the prevention of severe injuries in crashes involving large trucks.

Piwi-interacting RNA (piRNA) is a newly identified class of small non-coding RNAs. It can combine with PIWI proteins to regulate the transcriptional gene silencing process, heterochromatin modifications, and to maintain germline and stem cell function in animals. To better understand the function of piRNA, it is imperative to improve the accuracy of identifying piRNAs. In this study, the sequence information included the single nucleotide composition, and 16 dinucleotides compositions, six physicochemical properties in RNA, the position specificities of nucleotides both in N-terminal and C-terminal, and the proportions of the similar peptide sequence of both N-terminal and C-terminal in positive and negative samples, which were used to construct the feature vector. Then, the F-Score was applied to choose an optimal single type of features. By combining these selected features, we achieved the best results on the jackknife and the 5-fold cross-validation running 10 times based on the support vector machine algorithm. Moreover, we further evaluated the stability and robustness of our new method.

In this paper, the optimization issue of electric vehicle charging station layout (EVCSL) for tourist attractions is addressed, and an improved PSO is used to solve the optimization issue. Specifically, the improved particle swarm optimization (PSO) is proposed to obtain an appreciative planning solution of EVCSL, and dynamic weight adjustment strategy and integration into the two-population genetic mode are proposed to improve the optimization quality for PSO. Simulation results show that the proposed improvement strategies can increase the optimization quality for PSO effectively so that a more appreciative planning solution of EVCSL can be obtained.

Severe and fatal crashes involving large trucks result in significant social and economic losses for human society. Unfortunately, the notably low proportion of severe and fatal injury crashes involving large trucks creates an imbalance in crash data. Models trained on imbalanced crash data are likely to produce erroneous results. Therefore, there is a need to explore novel sampling approaches for imbalanced crash data, and it is crucial to determine the appropriate combination of a machine learning model, sampling approach, and ratio. This study introduces a novel cluster-based under-sampling technique, utilizing the k-prototypes clustering algorithm. After initial cluster-based under-sampling, the consolidated cluster-based under-sampled data set was further resampled using three different sampling approaches (i.e., adaptive synthetic sampling (ADASYN), NearMiss-2, and the synthetic minority oversampling technique + Tomek links (SMOTETomek)). Later, four machine learning models (logistic regression (LR), random forest (RF), gradient-boosted decision trees (GBDT), and the multi-layer perceptron (MLP) neural network) were trained and evaluated using the geometric mean (G-Mean) and area under the receiver operating characteristic curve (AUC) scores. The findings suggest that cluster-based under-sampling coupled with the investigated sampling approaches improve the performance of the machine learning models developed on crash data significantly. In addition, the GBDT model combined with ADASYN or SMOTETomek is likely to yield better predictions than any model combined with NearMiss-2. Regarding changes in sampling ratios, increasing the sampling ratio with ADASYN and SMOTETomek is likely to improve the performance of models up to a certain level, whereas with NearMiss-2, performance is likely to drop significantly beyond a specific point. These findings provide valuable insights for selecting optimal strategies for treating the class imbalance issue in crash data.