Search statement

The results presented by Web search engines are generally listed in rank order of most to least relevant. Algorithms used to establish relevance vary. Traditionally relevance has been defined by some measure of content agreement between the search statement and the results however, some of the Web search engines now use measures of site popularity (number of links or number of times accessed) to rank site relevance. Link analysis does temper the artificially high relevance created by sites that optimize their relevance with keywords, multiple titles, and other techniques [50]. However, valuable sites with fewer links or less general interest may be more difficult to find. 

Frequently one will find new citations (not found in the database just completed) in the next bibliographical database that one searches. Are these new citations unique to this second bibliographic database, or were they present in the first database but your search did not find them To build quality control into your search process, return to the first bibliographical database and search for the newfound citations using an author or a title-word search statement. If you do find the newfound citations in the first database, explore the citations index/key words. You may discover additional appropriate search terms or procedures to improve your search. 

An excellent source for definitions of these terms may be found at http // www.google.com/. In the search statement, use the syntax definerxyz where xyz is the term for which a definition is sought. 

Google Search is a popular search engine and is an essential tool. Google provides specific search terms to narrow down and refine a search statement. Refer to Appendix W for Google Related Search Criteria (Google Insights) . 

The NOT operator excludes single search terms from retrieval. But mind its hidden dangers. If the search term is e.g. methyl cellulose and you exclude terms like ethyl cellulose, alkyl cellulose, etc., relevant documents might be forgone, where both methyl cellulose as well as other cellulose ethers are contained. Therefor you should be extremely careful when using the NOT operator. It is advisable to refine the search statement by further AND combinations. In any case, the amount of data to be excluded should be dealt with prior to other operations in order to check via titles, etc. if those documents should really be excluded by the NOT-operator. 

Operators have their hierarchy If you require more than one operator for one search statement the NOT- and AND-operator receives highest priority at STN, followed by the OR-operator. Every host has its own hierarchy of logical operators. If more operators are to be combined for one search, use always brackets, since bracketed commands are dealt with first. 

The words decanedioic and acid must be adjacent. Additionally, the word ethylhexyl must be part of the same chemical name. Search statements may become very long, as shown in the following example for 1,2,7,8 tetrachloro-dibenzo-para-dioxin ... 

The non-specific derivatives e.g. include polychlorinated biphenyls or dioxi-nes. They cannot be found without the role indicator D being added to the Registry Number. This search statement will also retrieve all the documents with a role indicator DP added to their Registry Number. 

A reader of printed matter automatically corrects spelling mistakes without thinking while reading. During an online query, however, the computer is not flexible and will only find words written in the predefined way. On the one hand, mistakes should not occur when writing search statements, but on the other, correct spelling in the search statement will exclude terms that are incorrectly written. The next example will show that databases are not free of faults ... 

Delete Last Search Statement. .PG DELETE. DELETE REMOVE. .ER DEL LAST... 

In Search Statement 2, the CAS Registry Number and a few of the synonymous terms from CHEMLINE are ORed (combined) as search parameters. This results in the retrieval of 257 records containing the desired chemical substance. Thus, the CHEMLINE information effectively Increases the TOXLINE retrieval by 97 records. Search Statement 3 is designed to select those citations published in 1977 in which both Leptophos and some variation of the term nerve (nerve, nerves, nervous, etc.) appear. 

The intersection of those records retrieved in Search Statements 2 and 3 should satisfy the criteria established for the substructure search (6 records). 

After accessing the RTECS file (Figure 11), the name fragments common to each name listed in the CHEMLINE search (DIBENZO and DIOXIN) and the formula fragment CL are used to retrieve six records. The RTECS Toxicity Data fields (IX) are then searched for the terms ORAL, RAT and TERATOGENIC and the records containing these terms in the index string are Intersected with Search Statement 1 to retrieve two records (Search Statement 2). 

---