現今攜帶型裝置與行動網路相當普及，網路設備商蒐集到了大量用戶去識別化後造訪網站或是APP程式服務連線到主機端時的DNS記錄，儘管擁有如此大量的資料卻缺乏有效的分析方式。 本次研究將會以前綴投影模式增長法(PrefixSpan)來尋找行動網路用戶是否有特定的網路使用習性與時序，期盼能夠作為企業未來發展方針的參考依據。 實驗結果顯示行動網路用戶在這一周的資料當中平日與假日的Pattern並無太大差異，多以使用娛樂類型的服務為主，尤其音樂與影視方面有壓倒性的多數，而在序列方面使用者大多會在使用某類型的服務後繼續使用具有相同類型的服務，而不是轉為使用其他類型的服務。最後則是通訊軟體服務的使用頻率遠低於預期且智慧家電服務的使用頻率非常的高，甚至高過部分娛樂類型的服務。;Portable devices and mobile internet have become extremely prevalent today. Network device providers have collected a large amount of DNS records, which are anonymized, capturing user visits to websites or connections to server endpoints through various app services. Despite having such a vast amount of data, there is a lack of effective analysis methods. This research utilizes the PrefixSpan algorithm to investigate whether mobile internet users exhibit specific usage patterns and temporal sequences. The aim is to provide valuable insights as a reference for the future development strategies of companies. The experimental results indicate that there is not much difference in patterns between weekdays and weekends among mobile internet users in the data collected for this week. Users predominantly engage in entertainment services, particularly in music and video content. In terms of sequential behavior, users tend to continue using services of the same type after using a particular service, rather than switching to other types. Surprisingly, the usage frequency of communication software services is lower than expected, while the usage frequency of smart home services is exceptionally high, even surpassing certain entertainment services.
