Searching for the development trend of close-range peek meta-search

Intermediary transaction http://www.aliyun.com/zixun/aggregation/6858.html ">seo diagnose Taobao guest cloud host technology Hall

The current domestic things about the yuan search for two years to talk about less, recently because some of the reasons for concern to the domestic search engine, inadvertently turned to the information on the meta search, simply put the relevant things sorted out to share.

Just say meta search from the concept of more close to the needs of users, after all, 1+1 is not necessarily greater than 2, but more than 1 is certain.

What is a meta search:

Meta search engine (Meta-search Engine) is a search engine that collects, filters, deletes and optimizes the search results of multiple search engines.

Relative to the meta search engine, the independent search engine that can be used is called "Source Search Engine" (Source-search Engine), or "Member Search Engine" (Component-search Engine).

From the functional point of view, Meta search engine is like a filter channel: The output of a number of independent search engines as input, after some extraction, culling, extraction and other operations to form the final results, and then output the final results to the user.

Run process for meta search:

Users through a unified query interface input query request, Meta search engine to the query for a certain preprocessing.

Meta search engine according to the member search engine scheduling mechanism, select a number of member search engines.

Meta search engine According to the selection of the member Search engine query format, the original query please localize processing, conversion to the member search engine requirements of the query format string.

Send a formatted query request to each member search engine, waiting for the return result.

Collect return results from individual search engines.

The return results are integrated, for example, to eliminate duplicate links, dead links, etc., to form the final result.

Returns the final result to the user in a certain format.

Meta Search features:

There is no independent web database.

Can be based on the needs of users to synthesize the search results of specific search engines.

Ability to index specific types of searches, such as pictures, documents, Web pages, videos, and so on.

Meta-Search trends:

The research and development of meta search engine needs the theory and technology of information retrieval, artificial intelligence, database, data mining, natural language understanding and so on, which is comprehensive and challenging.

An ideal meta search engine should have the following functional requirements:

Cover more search resources, can choose and call independent search engines, but also according to a certain scheduling strategy for automatic scheduling.

With as many optional functions as possible, such as resource type (website, Web page, news, software, FTP, MP3, Flash, image, film, etc.) selection, wait Time control, return results quantity control, result time selection, filtering function selection, result display mode selection, etc.

Powerful retrieval request processing functions such as supporting logical matching retrieval, phrase retrieval, natural language retrieval, and other search engines to retrieve grammatical rules, character conversion functions (such as the search engine that does not support "NEAR" operators, can automatically be implemented by the "NEAR" to "and" operator conversion, etc.).

Detailed and comprehensive information description of the search results (such as Web page name, URL, Digest, source search engine, results and user search requirements of relevance, etc.).

Support for multiple language searches, such as Chinese and English search.

The results can be automatically classified, such as according to Domain name, country, resource type, area and so on sorting.

Personalized service can be provided for different users.

Based on the existing independent search engine, an efficient meta search engine can expand the processing ability of the independent search engine, improve the recall of retrieval, and possibly further improve the precision.

But the autonomy of each member search engine causes the difficulty of integration, the difficulty mainly comes from: The difference of retrieval interface, document indexing method, correlation function difference, query parameter difference, retrieval function strength and so on. The evaluation mechanism of search engine retrieval effect; The automatic scheduling mechanism of the member search engine; Design the method of the search engine description file, make the system have good expansibility, give its own result fusion algorithm, can trace the user's use, accept the user feedback to study and adjust, make the system have the self-adaptive.

Read the above introduction is not for the meta search has a certain understanding, domestic currently do a relatively good yuan search only compare cat, integrated including Baidu, Google, Yahoo, Sogou, search, Youdao, live, including numerous Chinese search engine; foreign meta search engine I recommend you can look at Dogpile, Clusty.

XJP from the original: http://www.xjp.cc/2009/04/16/meta-search-engine-guide/