OUTLINE Basic ideas of traditional retrieval systems image

OUTLINE • Basic ideas of traditional retrieval systems • image retrieval systems • differences

the architecture of information retrieval system 1． spider 2． database 3． User interface Spider is running on server, it can be closed and started. Spider parses the pages and segments the info stream and saves them into database. Database stores the abstract, date and URLs. The user interface is bridge between the user and database.

Search program 1）read the page stream based on the initial URL; 2）exact certain info from pages(including URLs), stores them into database； 3）repeats until no new URLs in queue, or time out ；

User interface User Interface retrieves the records from databased on key words.

AIL system: Simple Search

Advanced Search search choices: keyword, author, subject, size and modified dates

Result： abstract, whole text, size, date and result ranking

Keywords Relationship *： prefix “AND”，“AND NOT”： “OR”：

User’s templates User account, PASSWORD，EMAIL，subject and start URL.

Add, insert and delete templates

User template search user account, password, keyword, author, date, size

The shortcoming of traditional retrieval system： l. Traditional feature vector based retrieval method can not provide retrieval on the semantic level. l. Some spiders cannot even understand frames. l. Retrieval systems cannot yet make heads or tails of any images, audio or video clips, so these bits of information are wasted.

Image retrieval system The components of image retrieval system include: the user-interface network handling server programs

The user interface collects the query from the user and send the query string to the Server as parameters The communication across the network from user to server will be handled entirely by the HTTP protocol. Server will receive the cgi-bin URL parameters and facilitate the execution of the remaining server programs.

create query string

search result

The differences between the two systems image retrieval system is better than traditional retrieval system: l. The indexing part of context based image retrieval system is more powerful. Because images retrieve needs more requirements. l. The context based image retrieval system can do with the uncompleted and fuzzy queries. l. The context based image retrieval system can refine the results.

image retrieval system should implement the following points: l. Adopt the template conception into the image retrieve l. Combined with the traditional retrieval system l. More resources