Search results

The set-union knapsack problem is one of the most significant generalizations of the Non-deterministic Polynomial (NP)-hard 0-1 knapsack problem in combinatorial optimization, which has rich application scenarios. Although some researchers performed effective algorithms on normal-sized instances, the authors found these methods deteriorated rapidly as the scale became larger. Therefore, the authors design an efficient yet effective algorithm to solve this large-scale optimization problem, making it applicable to real-world cases under the era of big data.

The authors develop three targeted strategies and adjust them into the adaptive tabu search framework. Specifically, the dynamic item scoring tries to select proper items into the knapsack dynamically to enhance the intensification, while the age-guided perturbation places more emphasis on the diversification of the algorithm. The lightweight neighborhood updating simplifies the neighborhood operators to reduce the algorithm complexity distinctly as well as maintains potential solutions. The authors conduct comparative experiments against currently best solvers to show the performance of the proposed algorithm.

Statistical experiments show that the proposed algorithm can find 18 out of 24 better solutions than other algorithms. For the remaining six instances on which the competitor also achieves the same solutions, ours performs more stably due to its narrow gap between best and mean value. Besides, the convergence time is also verified efficiency against other algorithms.

The authors present the first implementation of heuristic algorithm for solving large-scale set-union knapsack problem and achieve the best results. Also, the authors provide the benchmarks on the website for the first time.

Hate speech is an expression of intense hatred. Twitter has become a popular analytical tool for the prediction and monitoring of abusive behaviors. Hate speech detection with social media data has witnessed special research attention in recent studies, hence, the need to design a generic metadata architecture and efficient feature extraction technique to enhance hate speech detection.

This study proposes a hybrid embeddings enhanced with a topic inference method and an improved cuckoo search neural network for hate speech detection in Twitter data. The proposed method uses a hybrid embeddings technique that includes Term Frequency-Inverse Document Frequency (TF-IDF) for word-level feature extraction and Long Short Term Memory (LSTM) which is a variant of recurrent neural networks architecture for sentence-level feature extraction. The extracted features from the hybrid embeddings then serve as input into the improved cuckoo search neural network for the prediction of a tweet as hate speech, offensive language or neither.

The proposed method showed better results when tested on the collected Twitter datasets compared to other related methods. In order to validate the performances of the proposed method, t-test and post hoc multiple comparisons were used to compare the significance and means of the proposed method with other related methods for hate speech detection. Furthermore, Paired Sample t-Test was also conducted to validate the performances of the proposed method with other related methods.

Finally, the evaluation results showed that the proposed method outperforms other related methods with mean F1-score of 91.3.

The main novelty of this study is the use of an automatic topic spotting measure based on naïve Bayes model to improve features representation.