Personalized travel recommendation has attracted lots of research attention in both academic and industry communities. Although a great progress has been achieved so far, existing travel recommender systems have not well-exploited users' style-oriented preference on landmarks and local preference on the targeted city. Typically, users have their own preferences on the styles of landmarks (e.g., natural scenes or historic sites). When visiting a city, their preferences will be affected by the characteristics of this city (e.g., historic or scenic) or the "must-go" landmarks, as well as local contexts such as distance and time constraints. In this paper, we propose a novel style-oriented recommender system, which considers all the above factors to facilitate personalized landmark recommendation. Specifically, we first propose a unified classifier to detect landmark styles based on domain adaptation by leveraging web-photos in the source domain and landmark-image in the target domain. The detected landmark styles are then utilized to learn users' style-oriented preferences based on users' travel records in the past. Next, given a targeted city, the influence of users' landmark style preferences and the characteristics of the must-go landmarks of this city are simultaneously considered by a proposed style-oriented recommender system to make optimal recommendations. In addition, we further study the effects of local contexts, such as landmark popularity or location, on the performance of landmark recommendation. Extensive experiments on the real-world travel data of six cities demonstrate the effectiveness of the proposed style-oriented landmark recommendation strategy.
The research of sources number detection is still open and challenging issue in array signal processing. The accurate estimation may be very essential to those high resolution direction finding algorithms. Three crucial issues are discussed for the application of cluster method to the source-number detection. A detection method based on Fuzzy-c-Means clustering algorithm has been proposed, which uses canonical correlation coefficients of the joint covariance matrix as the feature to be classified. Compared with the classical methods, our algorithm has better performance in low SNR and angular resolution. In the spatially correlated noise fields, the simulation results demonstrate the effectiveness and robustness of the proposed scheme.
Considering the unobserved spatial heterogeneity, this study aims to build a Bayesian spatial multinomial logistic (BSMNL) model by utilizing the geographic information from historical maritime accidents. The proposed BSMNL model can be applied to investigate the determinants of human errors involved in maritime accidents. Compared to the traditional multinomial logistic (MNL) model, the proposed BSMNL model produces a more accurate estimate of the effects of environmental and accident factors on the occurrence likelihood of human errors in maritime accidents. Results show that accidents involving cargo and container ships; tankers carrying liquefied natural gas (LNG), liquefied petroleum gas (LPG), or oil; and fishing vessels are more likely to be associated with human errors. Further, one important finding is that the involvement of fishing vessels significantly increases the occurrence probability of both negligence errors and judgment or operational errors. In addition, the occurrence likelihood of human errors is generally higher in springtime, conditions of poor visibility, the absence of strong winds or waves, and the moored or docked status.
For effective maritime traffic emergency rescue (MTER) operations in the event of maritime traffic accidents (MTAs) and to improve rescue efficiency, it is necessary to analyze the MTER synergy problem and the cooperation between port states. First, the spatial information of accidents under the geographic information system data structure is clarified from the global integrated shipping information system of the International Maritime Organization, and the density-based spatial clustering of applications with noise algorithm is used to conduct hotspot mapping analysis of MTAs to establish the clustering and classification of accident characteristics in key areas. Second, the classification characteristics of accident samples are extracted based on spatial information, and the correlation attributes between MTA hotspots are analyzed. Furthermore, by introducing complex network measurement technology, a topological model of the MTER network is established considering the correlation of accident hotspots, and this model is combined with the sample data of MTAs in Southeast Asian waters from 1990 to 2022. Third, the MTER topological network model is quantitatively analyzed under the accident space of Southeast Asia, and the degree of correlation of traffic accidents in key areas is obtained to reveal the inevitable demand for MTER between regions. The results of the analysis show that there is a network correlation between inter-regional accident hotspots, and thus the degree of correlation between accident hotspots needs to be considered for MTER in key areas. Countries in densely connected regions would set up joint rescue exercises and consider rescue assistance between port country stakeholders, thus improving protection for accident emergency responses. The method of complex network topology based on spatial correlation between accident hotspots suggests a new approach for solving the MTER problem.
In the smart mariculture, batch testing of breeding traits is a key issue in the breeding of improved fish varieties. The body length (BL), body width (BW) and body area (BA) features of fish are important indicators. They are of great significance in breeding, feeding and classification. To accurately and intelligently obtain the morphological characteristic sizes of fish in actual scenes, data augmentation is first used to greatly expand the published fish dataset, thereby ensuring the robustness of the training model. Then, an improved U-net segmentation and measurement algorithm is proposed, which uses a dilated convolution with a dilation rate 2 and a convolution to partially replace the convolution in the original U-net. This operation can enlarge the partial convolution receptive field and achieve more accurate segmentation for large targets in the scene. Finally, a line fitting method based on the least squares method is proposed, which is combined with the body shape features of fish and can accurately measure the BL and BW of inclined fish. Experimental results show that the Mean Intersection over Union (mIoU) is 97.6% and the average relative error of the area is 0.69%. Compared with the unimproved U-net, the average relative error of the area is reduced to about half. Moreover, with the improved U-net and the line fitting method, the average relative error of BL and the average relative error of BW of inclined fish decrease to 0.37% and 0.61%, respectively.
For a long time, selecting specific events from a large number of auroral sequences and analyzing them has been a routine process for space physics experts to study space activities between the Sun and the Earth. As a common space event with greater impact on the magnetosphere and ionosphere, substorms are a hot topic in the space physical field. Therefore, the automated identification of substorm sequences is an ongoing task. At the same time, due to the particularity of substorm sequence data, people who do not belong to the field of space physics cannot label and identify them, which largely limits the application of artificial intelligence techniques in this field. Therefore, this paper proposes a progressive method for encoding visual information to automatically identify substorm sequences. Based on a small amount of substorm visual map labeled by manually, this method designs a stream to learn and predict these visual maps. Meanwhile, the predicted visual map will assist the network in extracting representative sequence features and finally getting the identification results of substorm sequences automatically. The series of experimental results show the effectiveness of the proposed method.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)