Search is ubiquitous, with its shadow on every Web site, browser, application program, and platform. Search was born in the era of Internet information explosion to meet the needs of users to quickly find the target message, the form can be very simple but the database and the implementation of technology has a special high requirements.
Innovation in search is not easy, but it is not impossible. This paper first analyzes the characteristics of the search user requirements, and introduces the form of search, and then discuss some possible ways to improve the search experience.
Understanding Search
First, about users
The purpose of the user search is obvious, in order to find the answer information. Users at this time may know what they want, such as searching for a download link to the 33 days of lovelorn, a situation known as a project search. Users may not know exactly what they want, such as the search for a good movie as a pastime, which is called exploratory search. The search behavior patterns of these two scenarios are not the same, but they also have a strong purpose: the user is not interested in playing the search tool itself, eager to immediately find the information and answers that meet their requirements.
Fast forward fast out, is an ideal form of search, but also the desire of users.
Second, the Search performance form
Search box
The search box has become a search flag. Search box is the user search the entrance, but also the question of the problem area, but also the search engine command execution area.
Search box color, shape and size, different sites have different styles.
Although the box is small, it is very learned. In order to explain to the user what type of keywords can be entered in the input box, the search box can use "placeholder text", such as the search box on the first page of watercress.
Figure 1. The search box on the first page of watercress uses placeholder text
Users are not always able to spell the search keywords correctly, at this time your search if you choose to do nothing, then very likely to be users think your product is not smart enough, not smart. Google identifies the correct spelling in red when it encounters a spelling error, and does not erase your right to go wrong:
Figure 2. Google search to automatically correct the key words in the box
Users not only spell errors easily, but also often can't think of what to enter specifically. If the search box can automatically fill in the input process, and provide associated search suggestions, you can avoid user input errors, but also save the user's spelling time (which is more needed on the phone side).
Figure 3. Taobao offers real-time search advice
Search results
The content and presentation of the search results are different depending on the platform and the different product types. The search on Flickr is a picture, and a search on Weibo can get microblogging information, and search the App Store for a list of apps.
The search results page is much more learned. To do Internet products, we often explore the "paradox of Busy users": Users are busy, there is no time to travel through your complex information castle. Therefore, the search results page is best to streamline design details and think twice about your search strategy. This involves the trade-off between precision and recall, which is the relationship between this long and the elimination. Gao Chazhun rate, and keyword matching degree is very high, you may miss a lot of valuable information, Gao Chaquan rate, the approximate results are pushed to the user forced them in a variety of good, bad information screening, time-consuming and laborious.
In the existing search technology environment, Gao Chaquan rate and high precision is difficult to balance, at this time you need to find out whether your users prefer part of the relevant search results or more like all relevant search results.
In addition, it is also a matter of consideration as to which strategy to use to further filter the search results. For example, in a video search site or application, you can filter the search results by restricting the country, type, and Language options. Search for images in Google, you can further filter the relevant search results by color, size, type, and other options.
Figure 4. Google Image Search can filter search results by color and size
The above mentioned search box and search results function form is more mature, it is easy for users to type search keywords, reduce the user's information filtering costs. The details of how many real-time search suggestions to make, how to use historical search advice, and how to filter your search results need to take into consideration factors such as the level of background database construction, the specific objectives of the product, and the system platform that hosts the product.