What’s the relationship between retrieval and recommendation system in website
in order to solve the problem of information overload, numerous tools and solutions were invented, so far.
when we open any portal will find the website will have a search, or search, its role is to help users quickly find the information we want, but we also see the recommended information related to let the user to choose when we open a message in the this page. However, the search and information recommendation function is to better improve the user experience of the website, do not know if you have not thought about it in the end there is no relationship between what
What is the recommendation system?
but, if visitors or users without clear demand? For example, when I wrote this text, looking for background music, but if I open a music website directly, like Shanghai MP3 channel, in the face of tens of thousands of albums, it’s hard to say I have a very clear direction, often at a loss, do not know where to start. At this moment, I still face a problem of information overload, the essence is a very serious very widespread information overload, but the search engine or retrieval system has been difficult to effectively help me directly. At this moment, I need is an automated tool that can start from the music historical data, from the huge music library for me to find those songs with my listening habits and preferences for my choice, this tool can be called recommendation system.
search function in early appeared, I believe we should all be very familiar with, so we have to focus on the introduction of the later recommended function, we must fully understand that the recommended features can be better both in direct relationship what advantages and disadvantages of each other direct.
search engine retrieval system or the historical background of the birth of the basic user is in the face of information overload, how to quickly find some of their needs. For example, the need to purchase a new server to meet user needs, what should we do? There is a HP agent that can be directly near the phone or in his shop in a circle, to find a suitable model. Or if we do not want to go out, you can open the HP or Dell server in the official website, find the product classification list of them, or directly in the search bar search server. In the two example, purchasing are in fact facing some information overload, but stores may be very small, find the corresponding models do not need additional help, on the site because there are many different models to choose from, visitors will need to rely on classified information or retrieval bar you need to find the models. Here are two examples of the core is not overload level, the key lies in the user node or its visitors have a clear demand, at least independent visitors can describe this kind of information.
The essential difference between