Network search engine and intelligent agent technology

Search Engine | network The rapid development and maturity of the Internet in the world, prompting the rapid expansion of information in all areas of society, for people to find, obtain

Useful information provides a rich source of information, but it also presents a challenge to the accurate positioning of information. The search for online resources is

One of the important contents of the Network Information Service, and modern people also put forward to the correctness and comprehensiveness of the information grasping more and more high

, therefore, the most urgent task is to develop superior performance network information retrieval tools.


1993, after the birth of the first batch of search engines, development has gradually become mature, along with the computer intelligent research

With the development of adaptive and learning characteristics, intelligent agent technology is also transitioning from the experimental phase to the practical application. Current

, search engine (engine) and intelligent agent (intelligent agent) have become the key of network information search.

Technology and core ideas.


1 development status of search engine technology


1.1 Search Engine Technology


At present, the Internet Information retrieval technology is the most important search engine technology, the search engine is actually a dedicated WWW service

, can also be said to be a class of web sites on the Internet, which is different from the general site, its main job is to collect the network

Hundreds of thousands of websites and web pages of information form a huge indexed database. Use of excellent search engines can reach things half

The effect of the work times. At present, there are about 3,000 kinds of search engines on the network, we are more familiar with the website of Chinese search engine.

Dot has Sina, Sohu, Yahoo, NetEase and Chinese excite etc.


Generally speaking, search engine mainly takes two kinds of ways to realize the retrieval of network information resources, one is to use classified topic catalogue

In the form of a tree-like classification of the site, the linked site must at least belong to one of the categories to form a similar book

Library directory, the classification of the topic directory, the user through the browse these directories to find what they need, using this

Search engines have Yahoo, Sohu and so on, because of the use of experts to induce and classify, for information navigation to bring

Great convenience, but this way in the classification and catalog collation requires a lot of manpower; the second is to use keyword matching method

, its processing object is mainly text, it can establish a large number of documents by word (word) to document Index library, on the basis of which,

When a user searches a Web page with a keyword, the system displays all the sites, pages, and news that contain the search term.

and other matching information. Keyword search can solve the Web page details of the search problem, as long as the user input keywords, the system through the spider

The spider robot automatically retrieves the selected range and automatically marks the retrieved information into the index database.

Matches the page in the checked range to get the results of the search.


The information retrieval techniques used by 1.2 search engines and their shortages


At present, the information retrieval technology used by search engine mainly includes: robot technology, indexing technology, translation technology, conversion technology, filtration technology, database technology, result processing technology and so on. The biggest advantage of search engine is: the coverage of information is large, information

Novel, and the relevance of the search results in the ranking, the search engine to its high relevance of the search results arranged in front.
But because the search engine uses the information retrieval technology intelligence level restriction, as well as to the natural language understanding restriction, to the network

There are many deficiencies in the retrieval of information. There are several main aspects.


(1) Now the search engine mainly through the robot of all or part of the content of the Web page download to the database, because many of the pages downloaded is useless or temporary information, both affect the speed of retrieval, but also increased the user's search burden.


(2) As search engines are generally used keyword search methods, but in many cases, it is difficult for users to simply use the keyword or

The group collocation between the key words accurately expresses the information content which is really needed, and the difficulty of the retrieval is caused by difficult expression.


(3) The coverage of each engine is quite limited. The study found that no one search engine index more than the entire

1/6 of the Web page.


(4) The results of the search are not accurate. The accuracy of the search results is determined by the relevance of the query words to the Web page, often entering

A single query Word can return tens of thousands of results, or 0 results.


2 Intelligent Agent Technology


2.1 Intelligent Agent


Intelligent agent is also called the intelligent Body, is the new achievement of artificial intelligence research, it is in the case that the user does not have explicit specific request, according to the user needs, replaces the user to carry on each kind of complex work, like the information inquiry, the screening and the management, and can speculate the user's intention, the independent formulation, the adjustment and the implementation Intelligent, is an advanced, complex automatic processing agent software. Intelligent agent can be applied to a wide range of fields, it is a hot topic in the field of artificial intelligence, it is applied to the field of information retrieval, and it becomes one of the important technologies of developing intelligent and personalized information retrieval.


2.2 Characteristics of Intelligent agent


① of intelligence. With a wealth of knowledge and a certain degree of reasoning ability, can speculate on the user's intentions, and can handle complex difficult tasks, the user's needs can be analyzed to receive, automatically reject some unreasonable or may bring harm to the user requirements, but also from the experience of continuous learning ability, appropriate self-regulation, improve the ability to deal with problems.

② of the agent. In the functional is a user of some kind of proxy, it can replace the user to complete some tasks, and the results of the active feedback to the user.

③ Mobility. Can roam to any target host on the network, and carry on the information processing operation on the target host, finally return the result set to the starting point, and can move with the computer user's movement.

④ Initiative. According to the user's needs and environment changes, the initiative to users to report and provide services.

⑤ of collaboration. Through various communication protocols and other agents to exchange information, and can coordinate with each other to complete complex tasks.

3 combination of search engine technology and intelligent agent technology

Search engine and intelligent agent technology have their own advantages and disadvantages, the combination of these two technologies will be to develop a new generation of more powerful online information search system to provide a vast world. The intelligent agent mainly integrates the client's special environment, completes the search with the user interest. It identifies, induces and summarizes the user's information needs and preferences, analyzes the interests and interests of users, and automatically and independently proxies users to find information of interest through learning good rules. Combining search engine with intelligent agent technology is the inevitable trend of establishing new retrieval mode.



The introduction of 3.1 server-side personalized service


In the server to absorb the idea of intelligent agent technology, the introduction of personalized and humanistic service ideas. The user feedback mechanism is introduced to perfect the retrieval mechanism, to improve the hit rate, and to provide special retrieval service for individuals. This way can be implemented through the account system, that is, for each user to provide an account (similar to personal mailbox), to record the user's query trail, so that when the user again log in conjunction with the previous user retrieval records to cooperate with the provision of related retrieval services.


This model embodies the characteristics of personalized service, for users to query information can be directly extracted from the user's database, to avoid duplication of inquiries. In addition, the user's feedback is tracked to obtain the user's evaluation of the results, thus improving the quality of the search. The natural language development of the retrieval entrance will help to optimize the retrieval interface and improve the humanization of the interface.