Fine analysis of user behavior
we often talk about the Internet in product data operation, data operation is the key of user behavior data, because only the mining of user behavior, can make to attract more users of products, also is to make the user behavior analysis. Now everyone is talking about big data analysis and data mining, in fact, also contains the user’s behavior analysis. The traditional analysis of user behavior, generally refers to PV, UV, bounce rate and user access depth, user retention time, in fact, these indicators is mainly used as a statistical, real to direct product and business is to have a more precise analysis of the user behavior is fine analysis. For example, we have made a kind of product information, users click rate is there, but how to know whether users love as a channel, this time on the need of multiple users in the residence time, frequency, review comments and content to do more sophisticated analysis, quality or simply from the click rate didn’t mean content. So, what are the aspects of user behavior analysis,
first of all, traditional user behavior analysis has evolved into fine analysis. The traditional user behavior analysis, we usually think so: as long as a user opens the APP even if this day one of the active users, today’s new users as long as second days to start a APP that is a user retention. Now, our analysis of user behavior can be even more refined. For example, we do an internet reading class APP products, if a user opens the APP and not to an article view, but generated by registering machines and other means of zombie users, so we can set a rule to define what is the active users, at least today see an article the user is an active user. For example, many users as long as the traditional database to generate a data can be used as a user, and now many Internet products require real name authentication, also is a customer if you do not meet the conditions, it is not a valid user name authentication, such as mobile phone, rather than social products the certification. Therefore, the fine analysis of user behavior will make more and more Internet products, especially those who rely on false data to get the investment company more and more disappointed, because the fine analysis of user behavior will make a product more and more healthy, more and more high quality.
secondly, fine analysis of user behavior focuses on user profiling. We do user behavior analysis, although different methods vary greatly, but the ultimate goal is the same, that is, through the user behavior data generated to restore the user’s true behavior and the true face. That is to say behind user behavior in Internet products and the behavior of the user the time frequency based on the dimensions, as accurately as possible to restore the true user, through behavior data to construct a complete and precise user portrait, the features of data and user behavior data together constitute a portrait of the user instead, like the traditional use of statistical tools to get data.