Abstract:Stable environment is of vital importance in the process of long-distance circulation for agricultural products. With the development of Internet of Things (IoT) technique, it is relatively convenient to acquire the real-time data about the agricultural environment and the location data of the transport vehicle during the whole circulation process. In order to better perform the environment monitoring and position tracing, one can supervise the trajectory based on the collected IoT data through some current visualization approaches. However, the data collection of one specific trajectory on circulation for agricultural products is usually extremely large because the real-time data is usually required in real-world applications. Therefore, there still exist several open challenges to effectively and efficiently visualize the trajectory by large data collected on circulation for agricultural products. Firstly, there will be too many markers on the limited map once the visualization mapping is made directly from the original collected data set. Then the visual overload problem may occur when loading all the markers on the screen. Meanwhile, it also wastes computational resources to be rendered for the large scale data set, and it will decay the satisfaction of the end users. In addition, it is not conducive to grab the valuable information, which is typically usable for decision-making but hidden in the large raw data set. In order to effectively address these problems, a novel spatial-temporal visualization technique based on clustering the original data points is proposed in this paper. The clustering algorithm considers both the spatial-temporal characteristics and the semantic features of the data collected from the transport vehicle during the circulation process. For the spatial aspect, the trajectory is consecutive in nature, and the curves of the trajectory are well guaranteed with the constraint of the temporal factor. As for the spatial perspective, the nearby points are surely clustered together. Besides, the semantic features are taken into account, and then the points with abnormal IoT sensing values are detected in time, which is demanded by the manager of the circulation. In this approach, the original data will be firstly clustered by the proposed clustering algorithm and reduced to relatively fewer points, which are deemed to be critical positions in a specific trajectory on circulation for agricultural products. Then the real-time trajectory of this transport vehicle can be drawn with these critical positions to monitor the environment and trace its position during the circulation process. Furthermore, the proposed spatial-temporal visualization method is applied to the project that focuses on position-based analysis and visualization of agricultural IoT data based on the geographic information system. Finally, the system is successfully applied to several agricultural companies, and the intuitive visualization of the entire trajectory on the circulation process of agricultural products is effectively achieved.

DataTalk-V: time series visualisation for internet of things based on clustering and dimension reduction for an IoT platform

UDRN: Unified Dimensional Reduction Neural Network for Feature Selection and Feature Projection

Deep Clustering and Visualization for End-to-End High-Dimensional Data Analysis.

Using Visualization to Improve Clustering Analysis on Heterogeneous Information Network.

Using clustering algorithm to visualize spatial-temporal internet of things data in process of agricultural product circulation

A Visual Analytics Framework for Reviewing Multivariate Time-Series Data with Dimensionality Reduction

Subspace Clustering by Directly Solving Discriminative K-means

Clustering Time Series Utilizing A Dimension Hierarchical Decomposition Approach

DRSCDM: A Novel Density-Related Clustering for Complex High-Dimensional Data Streams

SOMTimeS: Self Organizing Maps for Time Series Clustering and its Application to Serious Illness Conversations

Comparison of Data Visualization, Outlier Detection and Data Dimensionality Reduction Methods

Topology-based Clusterwise Regression for User Segmentation and Demand Forecasting

Learning Linear Representation of Space Partitioning Trees Based on Unsupervised Kernel Dimension Reduction

Revisiting Dimensionality Reduction Techniques for Visual Cluster Analysis: An Empirical Study

IT-map: an Effective Nonlinear Dimensionality Reduction Method for Interactive Clustering.

KIHCDP: An Incremental Hierarchical Clustering Approach for IoT Data Using Dirichlet Process

A Novel Intelligent Clustering Approach For High Dimensional Data In A Big Data Environment

Computational intelligence based sustainable computing with classification model for big data visualization on map reduce environment

A Density-Center-Based Automatic Clustering Algorithm for IoT Data Analysis

RT-DBSCAN: Real-Time Parallel Clustering of Spatio-Temporal Data Using Spark-Streaming

Constructing a Prior-Dependent Graph for Data Clustering and Dimension Reduction in the Edge of AIoT