Research On The Spatio-temporal Fine-grained Organization Model And Portrait Application Of College Students’ Campus Activities

With the rapid development of the Internet of things and sensor technology,the construction of educational informatization has changed from digital campus to intelligent campus.At present,all kinds of Internet of things devices have been widely used in classroom,logistics,evaluation,scientific research,online learning and many other fields.These devices not only improve the efficiency of management,but also improve the level of teaching and management,and accumulate a huge amount of spatio-temporal data.However,the overall scale and complexity of these data are high,which brings difficulties in storage and query to deep-seated applications.Therefore,how to organize,manage and mine these low-value density spatio-temporal data with low cost and high efficiency has become an important content and direction of educational big data research.We have constructed the data organization model,established the efficient storage mechanism,and realized the deep application based on portrait.The main research contents are carried out from the following aspects:(1)We propose a spatio-temporal fine-grained data organization model for college students’ daily activities on campus,and put forward the data model and storage structure from the conceptual layer,logical layer and physical layer.First of all,on the conceptual level,a non-relational conceptual structure with students as the object is established,and the spatio-temporal attribute expression of daily activities is established on the basis of student objects.Second,the logic layer adopts the idea of object-oriented design,encapsulates students’ daily activities into three basic classes,and describes the object relationship of campus activities by means of inheritance and implementation.Finally,the physical layer adopts the design method of embedding MongoDB database,referencing documents and the idea of slicing design.We construct the mapping relationship between the student object and the storage organization,and realize the object-oriented organization and distributed storage structure.(2)We have constructed the spatio-temporal fine-grained student campus activity trajectory data set.The dataset includes multi-source data for 1845 undergraduates,including smart card consumption,library data,lending and WI-FI log data.First of all,aiming at the AP jitter phenomenon of WI-FI log data,we have used the shortest interval time and nearest neighbor location method to filter the WI-FI data,and achieved good results.Finally,we completed the fusion of multi-source data and obtained the campus activity trajectory data set including activity start time,residence time,location information and behavior description.(3)We have carried out feature mining based on spatio-temporal fine-grained activity.Data analysis is carried out from two aspects:the spatio-temporal characteristics of student activities and the spatio-temporal model of campus behavior activities.In the first aspect,the spatio-temporal characteristics are mainly from the three angles of time,space and activity types,combined with the comparison among majors,genders and grades to explore the differences in spatio-temporal characteristics.The second aspect,the spatio-temporal pattern mining of campus behavior activities mainly analyzes four aspects:social interaction,life regularity,residence degree and learning behavior investment degree.We also use the association rule Apriori algorithm to mine the strong association of college students’ behavior patterns.(4)We have constructed the portrait platform of college students’ campus behavior activities,established the label system,and realized the visual display of the portrait.The innovation of our research:(1)We have proposed a spatio-temporal fine-grained organization model of college students’ campus activities,and have realized the spatio-temporal serialization of finegrained data.(2)Based on the student object activity,we have constructed the non-relational organization method and storage structure,and realized the efficient query application of spatio-temporal topic.(3)Based on the space-time characteristics of activities,we have established a college student portrait analysis platform,and have realized the detailed analysis of individual and group behavior.