Machine search engine vs human flesh search engine (Author: Hu baojie)

Machine search engine vs human flesh search engine Hu baojie published on 16:29:13

More and more attention has been paid to social search recently. The most terrible social search is human flesh search.

The human flesh search engine, I directly put it into Q & A in social search, is actually asking a question in a community, A search mechanism in which a search engine is manually involved in the answer, rather than using machine-based Automatic algorithms to obtain results. From this perspective, Sina love question, Baidu know, Yahoo knowledge Hall are essentially human flesh search engines. Obviously, human flesh search engines are more targeted to extract search engine results through manual participation. Theoretically, they should better meet the needs of searchers than natural search results, especially for gossip search and multi-choice search without the only answer, the results of human flesh search are more attractive.

However, I personally don't catch a cold when using the search engine like Renren. Baidu knows that, for example, my book suddenly shuts down and is in urgent need of help. Click "I want to raise a question ", the feedback "sorry, please log in first"-the original question was to register first! If you ask a question, you have to wait and log on again. Some people may not have answered the question ...... Dizzy, no efficiency at all. If we judge that there is no market for human flesh, it will be wrong. The human flesh search engine is not designed for people like me who haven't even played in the community, people are prepared for those who love to participate, actively communicate, and find things. So we can see that the introduction of things such as the knowledge Hall is still a bit hot.

Human flesh search engines become popular. How can machine search engines respond? Only by analyzing user behavior in a deeper manner can machines understand user needs and provide feedback on results that best meet the search needs. That is to say, Google will continue to track users' click behaviors in the search engine results and browsing behaviors after clicking, rather than simply giving feedback to the results or tracing until the clicking ends. Because a link that does not meet the requirements (such as a cheating link), the user may exit quickly even if he clicks in. Theoretically, if the content of a linked page meets the search requirements, the user should stay on the page for a long time, or even walk around the site, and then generate more pageviews, this page may even be added to the favorites folder. However, if the search engine collects statistics on these user behaviors, will the ranking result be better as a supplement to the adjustment of pagerpank and various algorithms on the website's internal elements? No matter how subtle a human flesh mechanism it is.

Now it seems that Baidu is the best machine search engine in dealing with human flesh search engines (not because Baidu has a lot of manual intervention results, but Baidu is born with a human flesh search :), baidu thinking is not the Google-style technical thinking I mentioned above. What is "deep tracking and analysis of user behavior cloud ". Baidu's thinking directly refers to human thinking: by making a human flesh, and then organically integrating the human flesh results into the machine results, we can solve this problem perfectly. Of course, this is also true for other search engines, but Baidu's popularity is too high to be highlighted.

I still like and care more about machine search engines. Machines mean intelligence and efficiency. Although artificial intelligence and hard work are less than human search, search engines respect artificial intelligence, otherwise, Google and MSN will not use manually edited directories like dmoz as index data sources.