Research on Multimodal Data Fusion Strategy in the Construction of Personalized Service System of Digital Resource Library

College libraries are important information centers in colleges and universities, which are important positions for knowledge storage, communication and discovery.With the development of social digitization and informatization, the quantity and variety of information are changing day by day, and the information needs of university user groups are also changing.Research from data collection and analysis, label extraction to portrait mining applicable to the library user portrait model, combined with hybrid recommendation algorithms from the data collection layer, processing layer, fusion layer, service application layer four levels to build a multimodal dataenabled smart library structure.The K-means algorithm was used to cluster and analyze library users, and four types of users were formed: "pragmatic", "youthful", "recreational" and "curious".The accuracy, recall and comprehensive value F, which are commonly used to evaluate the recommendation effect, are chosen to validate the recommendation results, and the hybrid recommendation integrating user profiles is higher than the traditional collaborative filtering recommendation in recall and F value, and the recommendation effect is good.Index Terms user portrait, personalized recommendation service, K-means, smart library I. IntroductionIn the context of the information age, the way people acquire knowledge has also undergone new changes.The traditional library service mode has been difficult to meet the diversified and personalized needs of readers, who are no longer satisfied with simply borrowing books, but expect to obtain more accurate, efficient and convenient services [1]- [3].In this context, digital libraries have emerged, which provide readers with knowledge acquisition channels by virtue of rich digital resources, convenient access and wide coverage [4].However, with the increase in the number of digital library users, its management and service gradually tend to be complicated.How to effectively manage readers and meet their individualized needs has become an urgent problem for digital libraries [5]- [7].With the rapid development of the Internet, Internet of Things, social media and other technologies, huge and diverse data are constantly generated, which cover all aspects from individual behavior to social phenomena [8].The advent of the big data era implies the need to adopt new perspectives and technological means to understand, analyze, and utilize these massive data in order to better solve problems in the social, economic, and scientific fields [9]- [11].Digital library management supported by big data analytics technology emphasizes the efficient management, deep mining and intelligent application of data [12], [13].Through techniques such as advanced analytics, machine learning, and artificial intelligence, the laws, trends, and information hidden in multimodal data are discovered to provide more accurate support for carrying out personalized services [14]- [16].Therefore, in the era of big data, digital libraries are actively exploring personalized service models, striving to better meet users' needs, thus enhancing user experience [17], [18].This paper applies the user image to build the intelligent library field, starting from data collection, data processing, extracting labels to form the image and other aspects to gradually build the library user image model.Combined with the hybrid recommendation algorithm, the library personalized recommendation service system architecture is proposed on this basis, and a multimodal data-enabled smart library is constructed from four levels: data collection layer, processing layer, fusion layer, and service application layer.The collected data are processed to obtain different label types, and the K-means clustering method is used to achieve user classification, carry out personalized recommendation services for different users according to their characteristics, improve the utilization rate of library resources, and solve the problem of mismatch between book resources and readers' needs.