E-commerce website personalized search

SPAM online with more and more information on the Internet, people are surrounded by a huge database of the tide, the excessive information makes people get valuable content costs continue to increase, how to obtain more accurate information needed has become the issue that people need to solve. From Google and Baidu search in recent years changes, we can also notice how, for each user to create exclusive personalized search results interface, more and more become the direction of the development of search engine.

said that, I have to mention Google as early as 2008 has been released, the future of the search to rely on personalized view. At the same time the country’s largest search engine Baidu, also in 2011 released a new home page, the slogan is: Baidu new home, one person one world." Baidu based on user analysis, based on the algorithm and user behavior data analysis, can identify the user needs more accurate, more intelligent recommendation results page and related information to the user, so as to enhance the user conversion rate and enhanced user stickiness.

is based on the concept of user centered personalized recommendation not only changed the search engine, but also allows users with the same amount of information and e-commerce sites are aware of the heavy business opportunities. Through browsing time, shopping cart, purchase merchandise category and other behavior analysis of the content and user goods, electricity supplier website by screening interpretation of filtering data, and provide a better user experience for users, but also to achieve the purpose of sales promotion. As president Amason Geoff said: "if my website has one million customers, I should have one million stores." Recommended by personalized search technology, electricity providers do this site.

personalized recommendation algorithm in recent years, the electricity supplier has the following rules:

1, association rules: that is, to find the relevance of the sales records, so as to better guide the development of sales strategies. Typical case: "43% purchase Nestle Instant Coffee customers will buy Nestle coffee mate".

2, collaborative filtering: analysis of the target user to buy the goods, to recommend it and he had purchased goods similar to the goods. When defining the similarity of the commodity, either through behavior, two items with a user frequently purchased by whether at the same time, can also see through content, two commodity attributes or description is similar.

3, content analysis: the establishment of the user and the product configuration file, through the analysis of the content of the product has been purchased or browsed, the establishment or update the user profile. The system can compare the similarity between the user and the product configuration file, and recommend the most similar products to the user.

In fact, it is not difficult to imagine

, after the application of personalized recommendation on the site, when the user enters your site to search for target products, in the