基于神经网络word2vec的智慧物流比较研究

In the era of big data and artificial intelligence, intelligent logistics has become a hot topic of widespread concern in China. Word2vec is a neural network semantic model. First, this study collected data from China Federation of Logistics & Purchasing and China National Knowledge Infrastructure respectively. Corpora for media and academic platforms were developed based on the data and two word2vec models were trained. Then, based on the top 20 words which are most similar to intelligent logistics and logistics, comparisons of similar words were conducted. Following this, the t-SNE algorithm was utilized to reduce the dimensionality of the word vector. The matplotlib.pyplot package was used to visualize and compare the word embedding profiles of all words in the two word2vec models. The results manifested that there are differences in intelligent logistics between media and academic platforms. This study innovatively utilized the neural network word2vec algorithm to compare the intelligent logistics between media and academic platforms in China, providing a brand-new perspective for intelligent logistics research.