Research On Users’ Check-in Behaviors In Location-Based Social Networks

Location-Based Social Networks (LBSNs) is a new social network, and check-in is an important part of the LBSNs. Exploring behavior law of check-in can help businesses for precision marketing, and consumers can get better service from it, in addition, knowing the law is also important for design and development of LBS applications. On the context of this study, the author conducts a study about users’check-in behavior in LBSNs.The objectives of the research include:exploring the right tools and methods to collect LBSNs application data; exploring how to collate and analyze the data, in order to know the value of the results of the analysis.The author selects the typical location-based social network services website------Digu as research platform, the research process is divided into two stages:the first stage, through the literature review, the author finds that the third party tools GooSeeker can be more accurately for looking for target data, and can according to user specified data collection principle collecting Internet data than ordinary web crawlers, so the author uses the tools to collect more than 26,000 users and 1 million check-in information as the data source to MYSQL database. The second stage, the author does the research from check-in time, check-in places and check-in comments, exploring the method of processing and analyzing the data. In the second stage, the author focuses on the research in comment information selecting the comments containing "McDonald’s" as a study example, and uses text classification tool SVMCLS to divide into different levels of emotional tendencies, resulting in McDonald’s subjective preference. Then the author gets further analysis of the time distribution of the subjective preference and influencing factors.The main results of the study include:in the analysis of check-in time, the author finds users’ check-in periodic law, that is, one day in the morning check-in frequency is significantly lower than the afternoon, reaching a peak during the lunch hour, in the day of the week, Friday has higher frequency than other days; in the analysis of the check-in location, first-class cities have more check-in hot spots than other cities, the type of location takes consumption types as the core; in the analysis of check-in places comment, the author finds that users tend to give positive comments, and the time distribution has no significant fluctuation.In this paper, the author explores the tools and methods for collecting LBSMs application data, uses the proper methods in sorting out and analyzing the data, and summarizes the applied value of the results, the tools and methods can provide reference for the similar research.