Dnn search engine research

A few months ago, I studied dnn's search engine mechanism and principles. The reason for the study is that you want to use dnn's search engine, but find that it has very poor support for Chinese. To improve it and make it support Chinese, you can start to explore its principles. Since many dnn projects and source code are involved in the study, if you write out the principles of dnn through text, and make it clear to the readers, I am afraid it is a little difficult to express my words. Therefore, I made a PPT to try to get a better understanding of dnn search engines through graphic and visual forms.
The content of this PPT is personal research results. During this period, I referred to some online materials and the English documents of dnn3.2.2. I read the dnn-related source code in detail. If there are any mistakes, please correct me. At the end of the PPT, I proposed the idea of improving the dnn search engine. In fact, I have found a method that supports Chinese and English word segmentation, replacing the original word segmentation of dnn, it achieves the goal of enabling the search engine to support both Chinese and English texts, and has been put into use, with good results. If you are interested, I will consider writing Article .
In general, the key points of dnn search engines can be divided into three parts:
1. In the dnn architecture, an isearchable interface is provided. All modules that implement this interface can be used as search data sources. For example, the text/html module of dnn implements this interface, and the search result contains the text/html module content. Similarly, if you want to include your own modules in search engines, you need to implement the isearchable interface.
2. dnn uses the scheduling mechanism to split data sources. That is to say, you have added a text/html module to the page and entered an article. After the dnn search engine is scheduled to run, the content of this article can be searched by the search engine and will not be immediately displayed in the search results, because the scheduling is scheduled.
3. dnn has written a series of engineering projects to implement search engines. All projects related to search words, such as dotnetnuke. searchinput and dotnetnuke. searchresults. The PPT provided in this article mainly analyzes these engineering projects and does not have in-depth research on scheduling.
Note: After downloading the PPT, View in show Mode , The best effect, do not flip pages in the PowerPoint software, because I use a lot of dynamic effects to simulate the dnn search engine execution process.
Click to download: dnn search engine research