FAQ is just a question and answer!
A question-and-answer robot is a system that learns based on existing question-and-answer pairs, and then gives corresponding answers to users' questions.
Now is an era of knowledge explosion. Considering the efficiency of execution, machine learning is also based on the industry. It is necessary to provide the system with learning materials before it can provide relative industry capabilities.
- First look at the effect
Since I have been exposed to some operator-related things before, the existing FAQ is based on the operator's business FAQ. There are two ways to access,
1. Enter the text, return several related knowledge questions, and then the business system organizes and assembles the answers
2. Enter the text, and the FAQ system directly returns the business knowledge answer (the answer that needs to provide knowledge in advance)
Enter: http://1-blog.com/biz/nlpserver/faq?domain=operator&sentence=How to charge for mobile phone
Note that if you visit directly to see the effect, you need to use a browser with a url of utf-8 (for example, ie is not)
output:
Enter: http://1-blog.com/biz/nlpserver/faqanswer?domain=operator&sentence=How to charge for mobile phone
Output (of course the answer is wrong, how can I make these up):
In addition, if you are interested, you can try mobile reading, mobile newspaper, health guide, GPRS and other services.
- Original content for bots to learn
The original content is divided into two parts:
1. What is necessary is the questioning method of FAQ knowledge, that is, how you ask to indicate that you want this knowledge. For example, if you ask how to charge for a mobile phone report, it means that you want to know how much the monthly rent for a mobile phone report is. Provided The more you ask, the smarter the robot learns. The examples are as follows, which are the two knowledge questions about how much the mobile phone reports the monthly rent and how to check the attribution:
2. The answer to knowledge
It is the specific reply of each piece of knowledge, as follows
- Access method
First of all, I don't have so much industry knowledge. If you want to access this FAQ question answering robot system, you must first provide the questions and responses of the above example knowledge, and then you can access it. The system is developed for use, welcome friends who like it to contact me and access this question and answer robot system. PS, at present this robot system has no plans to do permission control, welcome to use~~~
Contact me: QQ 1319843502 