【摘要】 北大"天网"是一个大规模分布式搜索引擎系统.文中对其用户日志进行了分析和研究.结果显示:用户对系统的访问时间并不均等,一天中早晨、下午和晚上各出现一个波峰;通常用户在一天内只进行1-2次内容不同的查询,多于2／3的用户点击了结果页面中的某些网址(URL);多数用户输入的查询串中只含有一个词项并且包含中文字符,其中以2-4个汉字居多;用户查看结果页面的时间大约是2-3min;只有少数用户查看历史网页(或称网页快照).用户日志中不同查询串、不同用户和点击不同URL的数量满足Heaps定律.更多还原

【Abstract】 In this paper, the user log of Tianwang, a large-scale distributed Chinese search engine system, is investigated. The results show that the time distribution of users accessing the system is not uniform, and there are three peaks in one day, just in the morning, afternoon and evening. In general, only 1 -2 queries with different contents are carried out in one day by a user, and more than 2/3 of the users click some URLs in Web pages. Most queries contain only one word string with Chinese characters and the most frequent number of Chinese words is 2 - 4. The interval for the users viewing the result pages is about 2 - 3min, and only a few of users visited the historic Web pages (cached pages). Moreover, it is shown that the numbers of different queries, users and URLs follow Heap’s law.更多还原