February 28, 2012 
Site data analysis, time is one of the most common and indispensable dimension is used to limit the scope and granularity of the index statistics, in most cases, the time factor also caused some statistical rules and details of the indicators impact in some of the data analysis, we can easily ignore the factors of time, these effects may be misleading to the ultimate conclusion. Aware of this problem is a demand for data extraction, the site every day to release a lot of new content, new content recommendation, or will be buried, so many sites will have a "recommended" class module, This data needs is ...... »Read more: have to consider the time factor
October 27, 2011 
If you do the operators of the site's content, and recommend related content to help users quickly find and discover information of interest, so as to enhance website content browsing and smooth, thereby enhancing the value of the site conversion. Content is associated with recommended "and recommend to recommend the most common two associated recommendation is that we often say that the market basket analysis, namely the use of the purchase of a commodity purchase what this rule to find the goods between potential contact before the article describes - up marketing, cross marketing and the associated recommended; associated recommendation is recommended based on user behavior analysis, and the content recommendation is based on the content of solid ... »Read more recommended: KNN-based content May 10, 2011 
Optimize the website information architecture in the previous article in the blog - I have mentioned about the lost of the user (Lost Visits) definition, as well as how to use Google Analytics senior groups (Advanced Segment) to distinguish between these users. Measure (Measuring the the User Experience) looks at the "user experience", found that it is too tender, people Smith as early as in 1996 lost degrees (Lostness,) has defined, given the lost L formula Jiehuaxianfo, to others. Ie, L = sqrt [(N/S-1) 2 + (R/N-1) 2] L: Lost degrees N: Access ... »Read more: site lost measure March 28, 2011 
The most recently asked some users of statistical indicators, whether it is the decision-making departments, so this article focuses on some of the content that the user index. Suppose you want to concise data about the users of a site or product, you will be asked which user data? In fact a clever question is never how much of the cumulative number of users of the Q site, even the site of UV will not ask how much, because these indicators are not from the real meaning up to reflect the value of the site and development. Give a simple example - network Qin, the cumulative number of users should be no less than 10 million, but the figure is really able to reflect ... »Read more: it is of concern to user indicators March 13, 2011 
Described below is also at work a few days actually encountered and try to solve a problem. The company has a good team, everyone thinking about how to solve the existing problem of irrational, whether it is product design, content operations, or the background architecture. Below to solve the problem was a result of a data needs, according to the needs of design and summary of the thinking, but also need to constantly try to adjust and optimize. In fact, whether the commodity trading volume of e-commerce site, page views of content sites or resources website resources downloads, most are existence of this list. The charts provide a ...... »Read more: Top with a random number February 20, 2011 
In fact, the site analysis is very important one is the key to the site conversion path analysis may be many websites analysts in this regard are pouring a lot of time and effort to do our utmost to find the optimal path of conversion, because the optimization of the key conversion The path is equivalent to increasing conversion rates, and thus increase revenue. Therefore, especially for e-commerce site or a paid service sites, the key conversion path analysis is particularly important. Previous article - the site conversion rate and the definition and analysis of the funnel model on the critical path did a brief introduction, also recommend a very image of the data to show the way - the funnel model, can make every step of the conversion look .... .. »Read More: critical conversion path optimization January 30, 2011 
Has recently been very busy, so blog the update frequency will be relatively slow. Today, I want to talk about the data to show the aspects of the several views, variety of computing and processing data in the background, finally got some reasonable and intuitive indicators, these indicators we need to show to the demand side of the data, which will involve data presentation and data visualization. Is possible that these problems daily not data processing important data processing personnel are more concerned about data integrity, consistency, accuracy and timeliness of data show more data to reach the demand side of the hand where can rarely would anyone bother to focus on data ... »Read more: data statements and reports December 22, 2010
- The context of the data regardless of a site analysis tools, BI report or the report of the data, it is difficult to see the data in an isolated point alone, usually the data is the existence of sequence, grouping and other forms, for very simple reasons, we do not from a single data found for the analysis of the data must include the context (Context). The context of the data as one or some of the reference system for each indicator set by the reference and comparison process to analyze the pros and cons of data, like the example on the high school physics, if we do not ground as a reference, we can not distinguish between the train is stationary or marching, north ... »Read more: trend analysis of time series December 1, 2010 
First here is the title of the "User" refers to data users, or the demand side of the data, these users are often not the site or enterprise for external users, consumers of the data is usually the various departments within the company and personnel. Why would raise this issue, in fact, we often encounter such a situation: the company's executives complain useful things can not see from the report which is not can report finishing (so the following is busy open), but the how finishing or in the end they need the data (OK, the high level of demand is generally unspoken, we want to try themselves to try to figure out ... »Read more: What data is required? November 22, 2010 
Because recent changes have been busy with a hodgepodge of things, the transfer of work, termination procedures and the idea of finishing on the new job, is still in this stage, it may no recent new content to share with you several recent article is to summarize the contents of the main work before a pair of some thoughts. But believe richer data analysis relevant content goes to everyone, because I believe I'm going to the new company is a vibrant, full of creativity and challenge, and most importantly the importance of their data and understanding. See the title of the article, I believe we already know ... »Read more: the BI Application in the three major contradictions 
