**1. Introduction**

Along with the rise of a city network, which is constructed under the push of different kinds of urban elements flows, the interactions among different cities have been emphasized in the planning of urban areas, including the interaction of population, material, information, technique, etc. Hereinto, population interaction or population migration is one of the most important aspects. The floating of population is not only the flowing of individual human but also the transfer of demand, information, and technique carried by individuals [1,2]. They discriminately impact economic, social, and political development of both resettled areas and out-migrating areas [3,4]. Timely measuring and analyzing of population migration are particularly crucial for suitably planning urban space and distributing urban resources.

Related explorations on population migration have been concerned as hotspots since the 1990s. A larger body of researches have been conducted, such as the labor market performance, social and physical status of migration [5–7], the causes of migration flow [8–10], the consequent impacts of migration [11–13], the changing migration policies [14–16], the classification research of population migration [17,18], the spatial pattern of population migration [19], etc. These researches have been

conducted mainly based on three kinds of data: national censuses data, regional field survey data, and cyber big data. In the traditional migration researches, population censuses and field survey are the principal sources to provide population data [20,21]. For instance, Zhu [22] explored the determined factors in urban area which influence migrants' settlement intention based on the data from a survey on the floating population in the coastal area of Fujian Province. He et al. [23] adopted national census data to examine the distinctive spatial patterns of floating and Hukou population and evaluated their consequent impact on Chinese urbanization and industrialization. With the development of cyber space and the popularization of personal mobile termination, numerous researches have implemented under the assistance of data from cyber space exploration of the change, characteristic, and pattern of population migration [24–27]. For instance, Blumenstock [28] analyzed migration pattern based on mobile phone records and revealed more subtle patterns that were not detected in the government population survey. Zagheni et al. [29] used geolocated data for about 500,000 users of the social network website "Twitter" to predict turning points in migration trends and to improve the understanding of migrant populations.

Those researches have contributed largely to promoting the understanding of the progress of population migration and their impact. However, the deficiencies in migration data still exist. Studies based on national censuses data can explore the migrants in a large range but with a relatively large time interval of ten years, which hinders the short time-series analysis of population migration, and little can be inferred for specific years between censuses and for recent trends [29]. The researches based on field survey can provide detailed migration information, but it asks for a lot of time, manpower, and material resource to deploy, which are expensive for many researches. Simultaneously, the field survey often has a certain spatial location and cannot cover a large spatial scope. The increasing cyber data has opened up a new opportunity to deepen our understanding of population migration. However, studies based on the network big data always need to deal with extensive data and complicated procedures in acquiring and processing the data. At the same time, some data sources are not available openly, such as cellphone signal data and GPS data of resident activities, because those types of data include much individual private information that is protected by national law. A type of data with open, timelier, and easy-taking characteristics is necessary for effectively investigating the migration population.

With the growing application of search engine in cyber space, search query data has been brought out to reflect the preference of public attention, which is generated from the personal behavior of Internet search. This kind of data with opening and timelier characteristic has provided effective support for analyzing regional phenomena and problems [30–33]. In such context, the concern is triggered about its applicability in population migration research. In current information era, most people tend to take migration after an inquiry of destinations. Web search engine as the most widely used Internet tools provides massive information to the migrants and obtains relevant public attention on the specific subject of migration. The relationship between Internet search query data and population migration deserves more attention. However, the relationship between them is still unclear and there are a number of questions to be raised: can the search query data generated from migration-related information search offer some clues about population migration? If they can, how are they related? Do cites with higher cyber search quantity have a larger migration population than the cities with lower search quantity?

Based on these questions, this paper endeavors to answer them and to propose a new angle to analyze population migration. A hypothesis can be made that the search queries generated from individual migration-related search can positively reflect population migration. Based on the search query data from Baidu search engine, we construct a series of migration attention indexes (MAIs) to explore public attention on migration. Taking three main urban agglomeration areas of China as study area, the correlation analysis has been utilized to explore the relationship between MAIs and population migration to test the hypothesis. This paper is organized as follows. Section 2 introduces the study area and data. Section 3 elucidates the methodology of this paper, including the method and indicators that we applied in this paper. Section 4 reports the result of correlation analysis between

MAI and population migration. Section 5 conducts further discussion based on the results in our study area. Last, we conduct the conclusion of this paper.
