As one of the pillar industries of China's national economy, tourism is an important driving force for achieving consumption upgrading and economic transformation. With the arrival of the era of mass tourism, the large-scale and personalized tourism demand has put forward new requirements for the high-quality development of scenic spots. Therefore, studying the spatio-temporal behavior of tourists in scenic areas can provide a better understanding of tourists' preferences, which has certain practical significance. In the past, research on the spatio-temporal behavior of tourists mostly focused on large and medium-sized scales such as countries, cities, and regions, with less focus on the scale of scenic spots. And this study takes the Siguniang Mountain scenic area as the research object, based on the trajectory tourism digital footprints released by Siguniang Mountain tourists on the two-step road platform. The research studies the spatio-temporal behavior of tourists in the scenic area from a micro level perspective. Using Markov matrix to study tourist transfer, using kernel density analysis and average nearest neighbor analysis methods to study the spatial clustering phenomenon of tourist interest points and behavior trajectories, and studying the temporal changes of tourists through longitudinal time comparison. The results show that in terms of time characteristics, tourists' travel time shows a clear seasonal feature, with travel time concentrated in summer and greatly affected by legal holidays; In terms of spatial characteristics, tourists prefer Haizi Valley and Changping Valley, and their tourism routes roughly follow a linear pattern. Through the study of tourists' spatio-temporal behavior, the research results in the field of tourists' spatio-temporal behavior have been enriched, and suggestions and management basis have been proposed for the optimization and adjustment of tourism space in Siguniang Mountain.
With the development of artificial intelligence (AI) technology, a large number of AI products have been produced, which promotes the transformation of human-computer interaction (HCI) mode and puts forward higher requirements for HCI design. The application and development of HCI technology in AI products are systemically carded, the development trend of HCI technology in recent years is analyzed, and the three-dimensional architecture of HCI oriented to AI products is constructed, clarify the scope and boundaries of HCI design. This paper studies the necessity and feasibility of HCI design standardization, combs the standardization requirements, and combines the current standardization work of HCI design at home and abroad, creatively puts forward the standard system and key research contents of HCI design.
The objective of this study is to investigate user knowledge collaboration in open innovation communities, specifically focusing on the impact of knowledge exchange behavior on user knowledge output within user knowledge collaboration networks. [Methods] To achieve this objective, we utilize Python web scraping technology to collect user-related data and extract content data from the Huawei Pollen Club. We employ various software packages, including Networkx, NLTK, and Pandas, to perform tasks such as text mining, natural language processing, network construction, and calculation of feature variables relevant to user knowledge collaboration networks. By transforming user interactions and knowledge exchange behavior into measurable indicators, we develop a model to analyze how network characteristics influence user knowledge contribution through knowledge exchange. [Conclusion] The study found that different modes of knowledge exchange can improve the quality of user knowledge contribution, but compared to the breadth of knowledge exchange, the depth of knowledge exchange behavior conducted by users is more beneficial for improving user knowledge contribution. Knowledge exchange depth can increase the recombination of knowledge elements and enhance the user's knowledge system and quality of knowledge contribution, thus being more conducive to improving user knowledge contribution.
To improve the extraction efficiency and visualization of construction safety knowledge, this paper combines knowledge graph technology with construction safety domain, and proposes a basic framework of construction safety knowledge ontology based on the evolution logic of safety events according to the characteristics of knowledge. Considering two types of safety knowledge carriers and data sources, text and image, a knowledge graph is designed to contain text semantic features and image features, and the knowledge services based on different dimensional knowledge queries are validated in the experiments. The results show that the BERT-BiLSTM-CRF algorithm can be used to extract entities in text, and YOLOv5-FastPose can extract excavator poses from images. This paper verifies the applicability of knowledge graphs for safety knowledge mining, visualization and services.
To resolve the multifactor comprehensive assessment problem for the social vulnerability assessment, the projection pursuit dynamic cluster (PPDC) model combined projection pursuit principle with dynamic cluster rule is presented in this research, which is also adapted for social vulnerability assessment for the first time. The effectiveness of the PPDC model is validated by a case analysis of social vulnerability in St. Andrews Rural Municipality of Manitoba in Canada from published paper. The PPDC model can not only achieve the comprehensive assessment results and classification among the assessment samples at the same time, but also calculate every indictor weight according to the value of right projection direction optimized by the PPDC model. The studies show that the PPDC model is a powerful tool for multifactor comprehensive assessment, which is also an effective and new method for social vulnerability assessment.
