The application of interactive methods under swarm computing and artificial intelligence in image retrieval and personal
- PDF / 693,409 Bytes
- 10 Pages / 595.276 x 790.866 pts Page_size
- 71 Downloads / 144 Views
ORIGINAL ARTICLE
The application of interactive methods under swarm computing and artificial intelligence in image retrieval and personalized analysis Hangzhou Qu1 · Yinwei Wang1 Accepted: 2 October 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract The aim is to explore the interactive methods based on swarm computing and improve the image retrieval effects and personalized recommendation accuracy. The interactive methods based on swarm computing are explored. The mechanism of swarm intelligence (SI) algorithm is analyzed, in which the particle swarm optimization (PSO) algorithm and its improved algorithm are selected. The selected algorithms are combined with content-based image retrieval technology and applied to the image retrieval process, thereby realizing personalized analysis and recommendation based on users’ interests. Finally, the image retrieval behaviors of users are analyzed through simulation experiments, which verify the accuracy of the recommendation results. In the six sets of experiments, the image retrieval system based on the quantum behavior PSO (QPSO) has better performance compared to other PSO and SI evolution algorithms. The image retrieval accuracy of the proposed Bayesian personalized ranking (BPR) optimization algorithm (BPR-U2B) has significantly better performance compared to other recommendation algorithms. The QPSO algorithm is the best SI evolution algorithm for image retrieval. The BPRU2B algorithm is combined with the collaborative filtering algorithm based on BPR. It optimizes the objective function to limit the ranking results of the BPR algorithm, which is beneficial to complete the image recommendations and improve the personalized recommendation effects for users. Keywords Swarm computing · Artificial intelligence · Interactive methods · Image retrieval · Personalized analysis
1 Introduction With the development of information technology, the emergence of the Internet of things (IoT), and the continuous expansion in the application range of artificial intelligence, the emergence of IoT swarm makes the interaction and cooperation between the entities in different groups very important. Each entity has its autonomy. However, in the meantime, each entity will also work collaboratively for the benefit of the group community and may also interact with people connected to the network [1]. Currently, in the big data era, the birth of swarm computing is mainly to solve some complex computing tasks, thereby meeting a wider range of user needs. In particular, swarm intelligence (SI) derives from
B
Yinwei Wang [email protected] Hangzhou Qu [email protected]
1
School of Mechanical Engineering, Xijing University, Xi’an 710100, China
the background of artificial intelligence. As a more advanced performance beyond individual behaviors, SI is no longer a simple collection of multiple bodies; instead, it is presented by an emerging phenomenon of a group of individuals [2]. With the explosive and intelligent growth of data scale, the amount of real-time informat
Data Loading...
