Luo Chuan a list of statistical data, MySpace said Juyou Pageview the last six months increased by 338 percent, the user's coverage of the 138 percent growth, he believed that such a healthy pace of development, at least for now similar site does not have this type of development speed.
罗川列出一份统计数据，称MySpace聚友最近半年Pageview增长了338%，用户的覆盖率增长了138%，他认为这种发展速度属健康，至少目前同类网站没有这类发展速度。
For the first, we built a data preprocessing tool that verify the integrality and veracity; for the second a Pageview frequence-based, data independent, visitor's usage-center vector space model is proposed. Every Pageview is corresponding to a point in space. For the same Pageview denotes the different behavior sense in different session background, we use vector to describe user session. Then we cluster the user behavior vector to get the user class and the behavior sense vector which include the different behavior sense. In the end we can use the behavior sense knowledge to guide users.
本文首先根据Web使用挖掘的特点和要求，提出了一套高效准确的日志数据预处理方法和技术，以保证完整、准确、方便地实现日志数据的预处理；其次以页视图为基础提出了一种基于页视图同现率、独立于数据量、以访问者数据为中心的向量空间描述模型，使用向量描述会话，它可将会话向量聚类为用户行为类，从而得到代表不同的行为取向的行为语义向量，以指导访问者行为。
A pageview frequence-based, data independent, vistors usage-center vector space model is proposed. Every page view is corresponding to a point in space.
提出了一种基于页视图同现率、独立与数据量、以访问者数据为中心的向量空间描述模型，每个页视图对应向量空间中的一个点。
Uses your purchase and pageview histories to create a unique Web page that includes recommendations tuned to your taste.
比如Amazon.com根据你的购买和浏览记录来生成专为你的品味而推荐的图书。
For the same pageview denotes the different behavior sense in different session background, the vector is used to describe user session. Then the user behavior vector is clustered to get the user class and the behavior sense vector which include the different behavior sense.
由于页视图在不同的会话中含有的行为取向的语义信息不同，先将会话进行向量描述，然后将会话向量聚类为用户行为类，得到代表不同的行为取向的行为语义向量。