Mobile and Social Computing research group is affiliated with the School of Computer Science and Engineering, Beihang University.It has interdisciplinary research by combining computer technology with socialogy, mathematics, physics, communication technology and biology. The main focuses of the Research Group are mobile computing, social and information network analysis, complex networks, machine learning, image processing, data mining, compressed sensing, etc.The research group focuses on deep learning in various fields (such as medical diagnosis, smart grid, transportation networks, financial investment, image recognition, aviation management),as well as the research fields of complex networks, mobile handsets and social and information networks(the discovery and evolution of network community structure, the information and behavior of webcasting model), etc.
(A)Mobile Computing:
The Research Group has a wealth of experience in mobile phone project development. The mobile streaming media system based on smart phones developed by our research group, won gold prize of the Nokia smart mobile applications Challenge Cup in 2005. Since 2003, the group has carried out research and development of mobile phone software based on Symbian platform, launched software development based on the iOS platform for mobile phone in 2008, and carried out mobile software development based on Android platform in 2009.In the field of mobile computing,our primary study directions: mobile ad hoc networks, sensor networks, internet of things, indoor location technology, the video compression and reliability transmission technology based on the Internet, communication and image processing technique based on compressing sensing, fall detection technique and so on.
(B)Social Computing: The main research focuses are community discovery, network evolution, information and behavior spread, sampling algorithm for maintaining community structure and etc.In community discovery, the research group analyzed classical community discovery algorithms and proposed new algorithms.In the network evolution, the research group focuses on the assessment of the behavior of the nodes in the networks, the structure evolution of the connected components, and raising network evolution module. In information and behavior spread, we analyze the spread ranges and influence of information and behaviors in the networks, and further develop the spread module of information and behavior. The Research Group also researched thematic crawler, sentiment analysis and multi-document abstracts technology in order to meet the demand, and realized the automatic generation of field analysis report.Combined with social networks, this paper studies the proposed system's attack protection technology.In addition, the research group also conducted a bioinformatics research work on wheat pathology with a combination of complex network.
(C)Machine Learning: Based on machine learning, especially deep learning techniques, the research group conducted a research based on specific application requirements. Based on the medical needs, the automatic evaluation of bone age based on convolution neural network and heterogeneous learning, the intelligent diagnosis of deep malignant pulmonary nodules based on the fusion of heterogeneous features and the detection of eye diseases were studied.For target recognition, in combination with Generative Adversarial Networks, super-resolution reconstruction of images based on Transfer Learning and the generation of Generative Adversarial Networks was undertaken. Combining with the needs in the field of traffic, the 3D Convolutional Neural Network was used for traffic sign recognition. In the aspect of privacy protection, the privacy part of the image was encrypted based on the Generative Adversarial Sample technique. In the field of finance, we combined the machine learning method with old and new classification algorithms of banknotes and applied them to the sorter. We also applied deep learning techniques to stock index analysis and stock recommendation. In the field of aviation management, The research group predicted the Hard Landing of the aircraft and provided flight safety assurance. In addition, deep learning technology was used in the field of smart grid to predict electricity usage.
