Intelligent needs are reflected in two aspects With the development of mobile Internet entering a new direction, the intelligence in the mobile Internet has become a new development trend and major demand. The intelligent demand is currently mainly reflected in two aspects: On the one hand, it promotes new intelligent applications, such as autonomous vehicles, virtual reality and augmented reality applications, and broadens the mobile application field to provide users with more application choices. On the other hand, it is based on the existing large amount of application data for intelligent analysis, analyzing user needs based on existing mobile applications, clarifying user goals, providing user experience, and allowing users to experience improvements in the inherent mobile application field. In the stage of intelligent leading development, artificial intelligence technology is more and more widely used in the field of mobile Internet, and more and more artificial intelligence technologies are more involved in the development of mobile Internet. Because of its unique universality, autonomy and iterative optimization, artificial intelligence technology can deal with more complex data structures and data environments in data processing, resulting in more rigorous and stable models and derivation results. Artificial intelligence technology is constantly pushing the mobile Internet to complete new changes, complete more independent information capture, smarter analysis and judgment, more autonomous service delivery, and smarter cloud-to-end integration. This article will take the artificial intelligence technology as a starting point to further study the current situation of artificial intelligence solutions and applications in the field of mobile Internet. Artificial intelligence technology analysis Artificial intelligence technology has a wide range of research, including expert systems, neural networks, heuristic algorithms, fuzzy logic, genetic algorithms, and so on. At present, the widely used algorithms are mainly based on neural network, fuzzy logic and genetic algorithm. For example, the deep learning algorithm extended by neural network is the core algorithm of Baidu intelligent search. The artificial intelligence core technology can be divided into the following aspects when classified by application scenarios: Data mining and learning When faced with a large amount of data that requires deep data mining and clear communication between data, the commonly used method is an important branch of artificial intelligence - machine learning. Machine learning is the study of how to use computer simulation or to implement human learning activities. It is another important application field of artificial intelligence after the expert system. It is the fundamental way to make the computer intelligent, and it is also one of the core topics of artificial intelligence research. Its application extends to all fields of artificial intelligence. Learning is an important feature of human intelligence and a basic means of acquiring knowledge. Machine learning is also the fundamental way to make computers intelligent. Deep learning based on artificial neural network has been widely used. Neural network is an abstraction of some basic characteristics of human brain or natural neural network. It can realize functions by simulating some mechanisms and mechanisms of the brain. It is because of the multi-neurons, distributed computing performance, multi-layer deep feedback adjustment and other advantages of neural networks that it can calculate and analyze massive data, form models through data training, and its self-learning characteristics are very suitable for complex processing. Multidimensional nonlinear problems and massive searches based on intelligent associations. Knowledge and data intelligent processing The most used technology for knowledge processing is the expert system. The expert system is an important branch in the field of artificial intelligence research. It will explore the general thinking method and transfer it to the special problem to solve the special problem, and realize the major breakthrough of artificial intelligence from theoretical research to practical application. The expert system can be regarded as a kind of computer intelligent program system with specialized knowledge, which can apply the expertise and experience provided by experts in a specific field, and use the reasoning technology in artificial intelligence to solve and simulate various kinds of problems usually solved by experts. Complex issues. The key to developing an expert system is to express and apply expert knowledge. A basic expert system usually consists of a knowledge base, a database, an inference engine, an interpretation mechanism, a knowledge acquisition, and a user interface. As the earliest field of artificial intelligence, with the further development of artificial intelligence, the expert system has also promoted some new means, constantly updating existing research fields, and accomplishing mutual promotion with other artificial intelligence technologies. Human interaction The natural interaction and cooperation between humans and robots is to give robots the ability to observe, understand and generate various emotional features similar to human beings, enabling robots to interact like humans, and to perform functional assistance to meet the needs of human beings. . The main applications in human-computer interaction include robot learning and pattern recognition technology. The robot is a machine that simulates human behavior and is a relatively advanced technology in the field of intelligentization. The pattern recognition researched by artificial intelligence refers to the use of computers to replace humans or to help human perception patterns. The main research object is computer pattern recognition system, which is to enable computer systems to simulate the various perceptions of human beings through the sensory organs. . Mobile application performance analysis At present, the explosive growth of mobile application software has brought a large number of mobile applications with a large number of users, and there is no standard for users to select mobile applications. In this case, the friend-friendliness analysis and application software ranking of mobile applications have certain guiding functions for users to select mobile applications, and can also promote the benign development of mobile applications and improve the overall quality of mobile applications. At present, the performance analysis of mobile applications mainly focuses on end-to-end QoE analysis, user stickiness analysis, and business synergy analysis of mobile applications. The commonality of these aspects of analysis is that based on big data analysis, a large number of user and application related information is collected and calculated to draw conclusions. Then, in the process of big data analysis and calculation, artificial intelligence algorithms are usually used to solve. Different machine learning algorithms are powerful tools for big data analysis. Currently, there are enhanced learning algorithms and deep learning algorithms based on neural networks. The learning mode of these learning algorithms is to collect and store the energy consumption, delay, traffic, and rate of different applications in different application scenarios as the input end of the neural network system. Then, according to different application types, social applications, instant messaging applications, audio and video applications, cloud applications, browser applications, and game applications are differentiated, and applications in different fields are required for energy consumption, delay, traffic, and speed. The analysis is performed, and the weighted value of the corresponding indicator is generated as the weight of the neural network. Finally, a neural network model is formed by training and feedback calculation of a large amount of data. The learned model can complete the application's overall friendly conclusion output, and can also complete the application ranking based on the current data, while also predicting application performance through existing data. At present, each operator actively establishes a mobile Internet user behavior analysis system, analyzes the behavior of mobile users accessing the Internet on the basis of data sharing, and finally grasps the user's online habits and preferences, thereby accurately locating the user's demand for the mobile Internet. Provide data support for mobile enterprise business analysis and decision-making. Mobile app authentication Identity authentication is another important application direction for applying artificial intelligence algorithms. The most used authentication methods have always been password patterns for usernames and passwords, and devices such as U-Shields for peripherals. However, due to the possibility that the password is deciphered and forgotten, and U-Shield and other peripherals are inconvenient to save, the current multi-factor authentication method and biometric authentication method are being widely developed. The multi-factor authentication method uses two or more identity authentication methods to combine identity authentication, and combines the advantages of multiple methods to make up for the shortcomings of different identity authentication methods. At present, there is a multi-factor authentication that uses passwords combined with user big data map analysis for authentication analysis, and enhances the security of the original password login method. The user big data map records the data through various behavior patterns of the user in the Internet, performs analysis based on machine learning algorithms, completes the user's basic behavior model, predicts user behavior and rejects login or payment requests that do not conform to the user behavior model. Good assistance with password authentication mode. The biometric authentication method is an authentication method that uses a user's biological sample such as a fingerprint, a face, an iris, a finger vein, and the like to perform a comparative recognition. The biometric authentication method has high security, and with the continuous development of hardware, the biometric module carried on the mobile terminal is getting smaller and smaller, and it is more and more convenient to use. Because the number of biological information collection is huge, and the comparison database formed by feature extraction is huge, the algorithm basis of biometric authentication relies on artificial intelligence algorithm for pattern recognition and comparison authentication, and finally analyzes the recognition result output. . New application development Artificial intelligence technology has also spawned new applications and industries in the mobile Internet. Virtual reality/augmented reality technology is a new application based on artificial intelligence technology. Virtual Reality/Augmented Reality needs to quickly upload the server by collecting the perceptual data around the user, and deliver the result to the user's glasses device through server computing. For the user's experience, from data collection to result presentation, Google gives a latency threshold of 20ms. That is to say, in order to complete fast data exchange and calculation, in addition to the transmission bandwidth, there is a data calculation algorithm. The artificial intelligence algorithm realizes the instantaneous calculation of large data volume and solves the development foundation of virtual reality/augmented reality. In addition, a large number of applications relying on artificial intelligence technology (such as autonomous vehicles, smart homes, intelligent voice search, etc.) are rapidly developing. Baidu relies on artificial intelligence algorithms such as deep confidence network, convolutional neural network and recurrent neural network to solve applications such as natural speech processing, intelligent speech recognition and search, and image search. Artificial intelligence technology can have a huge impact on almost every area of ​​the mobile Internet. Data processing architecture based on artificial intelligence technology Based on the above research on artificial intelligence technology, a general data processing architecture based on artificial intelligence technology can be summarized. See Figure 1 for details. The data processing platform based on artificial intelligence technology generally includes a data acquisition module, a data training module and a data usage module. (1) The data acquisition module collects and stores all relevant data related to calculation, wherein the collected data includes historical data and real-time data. The data acquisition module is the data input of the entire data processing platform. (2) The data training module uses artificial intelligence technology to feedback and iteratively calculate the stored data of the platform, complete the data training work, and form a data processing model. The entire data training part is the core module of the data processing architecture based on artificial intelligence technology. The formation of the training model is closely related to the breadth and precision of the previous data acquisition, and the artificial intelligence algorithm used for training also plays a very important role. The formation of the training model greatly affects the results of the entire artificial intelligence data processing. (3) The final data usage module is based on the output of the training model. Generally, when the trained model is adopted, the prediction based on historical data and the calculation result of real-time data can be obtained, which is the output part of the entire artificial intelligence data processing platform. With the increasing range and number of calculations using artificial intelligence technology, the amount of computation based on artificial intelligence algorithm data processing models is also increasing. Usually, when performing artificial intelligence related operations, a lot of feedback and iterative calculations are performed, which will cause a large load on the server. When the amount of data that needs to be processed is further increased, the load of the data amount will slow down the performance of the server and affect the timeliness of the output of the result. Therefore, distributed computing is currently widely used to cooperate with artificial intelligence data processing. As shown in Figure 2, the distributed computing model can be distributed to multiple distributed servers for distribution. After the calculation is completed, the results are aggregated through different distributed servers and synthesized by the central controller. The distributed computing architecture and artificial intelligence computing complement each other to accomplish big data processing and computing tasks. Figure 2 Distributed computing structure Conclusion This paper studies the application of artificial intelligence technology in the development of mobile Internet, analyzes the core technology and application scenarios of artificial intelligence, and focuses on the artificial intelligence technology in the mobile Internet field for mobile application performance analysis, user identity authentication and the promotion of mobile Internet new applications. The important role played by the parties. In the context of the continuous breakthrough of artificial intelligence technology and the emergence of product innovation, this paper discusses the realization and development of artificial intelligence technology in the field of mobile Internet, and shows the penetration of artificial intelligence technology in various fields and the driving force of artificial intelligence technology in the industry. . Eleaf Vape pen, Eleaf wholesale, Eleaf disposable vape pen Shenzhen Xcool Vapor Technology Co.,Ltd , https://www.szxcoolvape.com
September 07, 2022