How far does graph search go?

Intermediary transaction http://www.aliyun.com/zixun/aggregation/6858.html ">seo diagnose Taobao guest cloud host technology Hall

When Facebook founder Mark Zuckerberg proposed a search, a lot of people expressed great interest, because Zuckerberg said that future search can be tailored, the temptation is very large. On the January 15, 2013, Facebook's search tool, graph Search, took the first step in a tailor-made search.

Tailored search, that is, everyone can according to their needs to get their most needed information, but also can be said to live with their own few people useful information. The foundation of all this is Facebook's strong user relationship. This user relationship is the lifeblood of the graph search organization's personalized results.

For example, search a "2010 High school classmate Party photos", the search of course is their high school class 2010 years of Class party photos, this search only for yourself or your class people useful, other people do not know how, and no interest. But by putting the search term in Google's search box, Google may return tens of thousands of question marks to you.

From this simple analysis, we can see some of the charm of social search, leaving a lot of vision and imagination, but also can find the challenges of social search. From these small examples, we can at least unearth two major challenges to social search:

First of all, the graph search wants to provide the user with the high quality information, needs the user the great cooperation. The greater the amount of user activity on Facebook, the greater the amount of data Facebook has, and the richer the user relationships, the more precise and fulfilling the results will be. For example, the above "2010 High school classmate Party photos", only someone to share the photos can be a picture, even if the photos are not there is no way to search, and the more photos uploaded, the richer the content. In addition, the more people share, the closer the relationship between students, the wider the scope of the spread. Also, the more information you have on each photo, such as name, background, subject, and so on, the more you can meet the search requirements and the better the content depth. And so on, the colorful things that users do around these photos, the more colorful the search results, the more valuable the search in the circle.

On the contrary, if the user activity in a circle is very small and the relationship is loose, it is hard for Facebook to organize the search in the circle. For example, the circle of people who see such a good thing is not "like", or "share", the graph search may not know who to recommend the content, may simply follow the traditional keyword matching to give the search results. The search results for keyword matching are very rough.

Therefore, to reflect the value of social search, the user's cooperation is very important. This kind of cooperation is only because of the one-sided needs of graph search, users will not be deliberately to cater to it because of its needs. This requires Facebook to have a good incentive mechanism to drive a variety of user activities within the platform, so that users are happy to build data and user relationships. This is a challenge, because user behavior is not mandatory, can only be cleverly guided, both to meet user needs and promote the search.

Second, social search faces another big challenge: changes in user habits. From the traditional search box to the social search box, it should be a completely different environment, and the two search boxes should host completely different search requests. Just like the "2010 High school classmate party photo", it is the default is to search in their own circle, the search is their own private things. And this search statement in Google's search box will not be satisfied with the search results.

In general, the traditional Google search is based on keyword search, a keyword behind may contain a variety of semantics, Google needs to speculate on this semantics and show the most popular results. But Facebook's social search is more biased towards natural language search, which is where it is competing with Google. Search requests are more in nature, closer to the natural language, which is the test of search techniques, but also to the traditional search habits of subversion. When Google and graph search two of searches did not separate the winners and losers, users need to switch between the two types of search thinking, which is a test for users, but also a chance for Facebook, because the search developed to the end, of course, everyone can use natural language to search for what they want.

Facebook's graph search is very beautiful and promising, but these two challenges are not to be overlooked. When people can do whatever they want, it's time for Facebook to solve both of these problems. At that time, Google, perhaps the search will decline, but it should also be in other areas to open up another one or even a few pieces of heaven and earth.

This article by Nanjing PLC Training http://www.njxxedu.cn/Original, reprint please indicate the source, thank you!

~

！