Literature Retrieval Data of Wenzhou Medical College in 2009

Information, knowledge and intelligence

Information: the latest reflection of the changes and characteristics of various things in the objective world, and the reappearance after transmission. Information is reflected by a certain form of material carrier, and it is the manifestation of existence state, motion form, motion law, their relationship and interaction. Reflect the different motion states and characteristics of everything, which can be divided into natural information, biological information, electronic information, social information and so on.

Knowledge: systematic theoretical information.

Intelligence: valuable information or knowledge that people transmit within a certain period of time for a certain purpose.

Literature: All carriers of knowledge are recorded.

First, the composition of literature:

Four elements: document information and document carrier symbol system recording method.

Second, the characteristics of medical literature:

1, a huge number

2. Diversification of carriers

3. Multilingualism

4. Interdisciplinary and decentralized publishing.

5. The updating speed of knowledge and information is accelerated.

6. The communication speed is accelerated.

7. Electronic development trends

Third, literature retrieval:

1 concept: the process of organizing and storing documents in a certain way, and finding out related documents or information content contained in related documents according to the needs of document users.

2 type:

1) Form of test result:

Bibliographic retrieval

fulltext retrieval

Reference retrieval

2) Information retrieval angle

document retrieval

Information retrieval

Fact retrieval

3) retrieval means

Manual search

Computer retrieval

3. Appearance features: it has also become a natural symbol, including title, author, source, issue, page number, year, month, type, number, language and other items.

4. Content characteristics: Also known as artificial marking, it refers to the topics discussed in the literature and the central contents of the literature, such as subject words, classification numbers, class names, abstracts, etc.

5. In the computer retrieval system, that is, the retrieval database, each document feature is called a field, the document feature identifier is the field value, and all the document features constitute a document title or abstract.

Fourth, the evaluation of literature retrieval system:

1, accuracy of reported information

2. Report information in time

3. Perfection of index system.

4. The depth of information indexing

5, recall rate and accuracy

Recall ratio (R) and precision ratio (P) are two most commonly used and important performance and effect evaluation indexes of retrieval system.

R= amount of relevant information detected/amount of relevant information in search tool X 100%.

P= relevant information detected/total information detected X 100%.

There is a reciprocal relationship between P and R. The higher R is, the lower P will be. Or conversely, Dallas to the audience can only improve relatively, and they will never reach 100%.

Five, the types of literature information:

1, according to the form of literature information:

Text type, video type, audio type and digital type.

2, literature information publishing type:

Books, periodicals, government publications, scientific and technological reports, patent documents, conference documents, dissertations, standard documents, product sample manuals, technical files, newspapers, press releases, statistical data, medical records, etc.

3, the carrier form of literature information:

Printed, miniature, audio-visual and electronic.

4, the processing degree of literature information:

Zero document information: unpublished or informal

Original document information: original data

Secondary literature information: retrieval tool

Third literature information: summary research and reference tools.

Six, electronic literature resources:

1 concept: It is a new file resource that stores various forms of files such as words, sounds, images and videos in the form of electronic data on non-printing paper carriers such as light and magnetism, transmits them in the form of electrical signals and optical signals, and reproduces them through external devices such as network communication, computers or terminals.

Two characteristics:

Changes in storage media and communication forms

The forms of information are diversified.

There are various types of literature resources.

Multi-level information service function

The update speed is fast and the timeliness is strong.

Have a retrieval system or function

Search access is not limited by time and space.

Types and products of medical electronic literature resources;

International common classification is divided into reference database, source database and mixed database. Among them, the reference database is divided into bibliographic databases (such as abstract index database and library catalogue database). ) and guide database; The source database is divided into full-text database, fact database, numerical database, terminology database and image database. Network electronic resources also include e-books, electronic periodicals, electronic newspapers, electronic news, academic websites, search engines/classification guides, subject navigation of network resources, FTP, BBS, discussion groups and so on.

Seven, literature retrieval language:

1 concept: a special recognition system used to describe the content characteristics and appearance characteristics of documents in the retrieval system and express users' retrieval problems. Express different concepts of things and reveal the logical relationship between concepts.

2 Logical relations: equivalence, subordination and association.

3 type:

1) Retrieves the controlled situation of words used in the language.

Standardized language: synonyms, synonyms, related words, polysemous words and abbreviations (to improve memory)

Non-standard language: free words, keywords

2) the time to retrieve the flag setting

Pre-assembly language: classification language, title language, etc.

Post-group language: narrative language, unit word language, keyword language, etc.

3) Information features described by retrieval language.

Describe the external characteristics of information: title of book/periodical/article, author/group author, citation, code/serial number.

Describe the characteristics of information content: classification language, code language and subject language.

8. How many theme languages are there? Briefly describe its characteristics.

Unit word language

Title word language

Keyword language: an nonstandard language

Narrative language

Nine, the basic methods, ways and steps of literature retrieval

The basic methods of 1 are: common method, retrospective method and subsection method.

2 retrieval ways: classification way, subject way, author way, numbering way, other ways (such as title, publication name, abstract, publication date, language, literature type, meeting place, etc. )

3 basic steps: analyze the subject and make clear the retrieval requirements.

Select retrieval tools and methods

Select the retrieval identifier and retrieval method.

Perform retrieval

Processing the retrieval result and obtaining the original document.

X. Methods and steps of computer literature retrieval:

Analyze the topic and determine the theme concept.

Select retrieval method and database

Determine the retrieval method: There are two kinds of commonly used retrieval words, one is subject words, and the other is free words. Keyword is a standardized retrieval word from the system thesaurus, and the use of keyword can ensure a high recall rate and accuracy.

Compilation of retrieval strategy (retrieval formula): 1) Boolean logic operator: ANDORNOT

The operator () is given priority, and the operation in parentheses has the highest priority.

Not> and> or

2) The distance between the words retrieved by the position operator in the record is only free words.

3) field qualifier

4) Cut-off operator: * and?

Feedback adjustment of retrieval strategy

Output of search results

1. recall and accuracy

Recall ratio (R) and precision ratio (P) are two most commonly used and important performance and effect evaluation indexes of retrieval system.

R= amount of relevant information detected/amount of relevant information in search tool X 100%.

P= relevant information detected/total information detected X 100%.

There is a reciprocal relationship between P and R. The higher R is, the lower P will be. Or conversely, Dallas to the audience can only improve relatively, and they will never reach 100%.

2. Retrieval language: a special identification system used to describe the content characteristics and appearance characteristics of documents in the retrieval system and express users' retrieval problems. Express different concepts of things and reveal the logical relationship between concepts.

Logical relations: equivalence, subordination and correlation.

3.MESH, (medical subject words), that is, the medical thesaurus.

The content mainly includes the change table of subject words, alphabetical order table, tree structure table and sub-topic table.

4. Compile the retrieval strategy (retrieval type):

Boolean logic operator: ANDORNOT, operator () takes precedence, and parentheses have the highest operational priority.

Not> and> or

"and", a and b, means that the document containing both search term A and search term B in the search results is a hit document.

"or" or "b" means that documents containing search term A or search term B or documents containing search terms A and B are all hit documents.

"No", where A is not B, means that the document containing the search term A but not the search term B in the search result is a hit document.

5.PubMed is a global free Internet-based biomedical literature database initiated by National Biotechnology Information Center (NCBI) under the National Library of Medicine (NLM), 1997. The database includes MEDLINE, citations in the process and citations provided by publishers.

6.PubMed retrieval operators and retrieval rules

1) Boolean logic budget symbol, not > and > or

2) Word-cutting symbol, where "*" represents zero or more characters. For example, chem* can be used to search documents containing words such as chemistry and chemistry.

3) Field identifiers, enclosed by "[]", indicate the search for specific fields, such as [ti] and [tiab].

4) Automatic conversion matching retrieval: PubMed first performs matching conversion on the input search words according to a certain thesaurus order, and then searches. The order of thesaurus matching transformation is

Net translation table, periodical translation table, introduction table and author index. If no matching words are found in the above four tables, PubMed will break up the phrases and repeat the above process in units of words or phrases. When searching, the logical relationship between each word is AND.

5) "",the phrase is retrieved as an inseparable whole.

South Dakota

Wildcard * means 0 or more characters? Represents a character.

Priority Or > W/n or Pre/n > and> nothing can change this order through brackets.

Proximity operator W/n (the interval between words and word levels does not exceed n words) PRE/n (the order before and after is required to be fixed twice)

Flue w/3 drug can detect documents containing fluent and drug, with no more than three words between them, and the order is not fixed.

When searching for a phrase consisting of two or more words, you must use quotation marks.

Author search first enter full name or abbreviation, then enter last name. The retrieval of the Authors field is different from the retrieval of the specific author field, that is, if the author field is limited, the search term can appear in different names; Restrict the specific author field, then the search term must appear in the same person's name.