Online searching: how to avoid the pitfalls.

How many times have you done a search and come up with hundreds of hits, or worse still no hits? Online databases are powerful tools for obtaining useful information, whether it is the latest statistics for paper production in China or a review of developments in impulse drying technology. Still, you need to know how to handle them. This article is a guide to avoiding some common pitfalls.

[ILLUSTRATION OMITTED]

When looking for pulp and paper sector information, it helps to use an industry specific database like Paperbase, available at www.paperbase.org. I will use syntax from this database in the examples below. Other databases may use other syntax.

Most readers know that search terms can be combined using the "Boolean operators" and, or, and not. If you were interested in the market for various packaging materials, you might search: market and (linerboard or "corrugating medium" or corrugated)

The search is therefore for records containing the word market and at least one of linerboard, corrugating medium or corrugated. The brackets indicate the order in which the operators are to be interpreted. Use quotes to retrieve only the two words indicated, adjacent and in the order specified.

The not operator requires more care. For example, if the searcher is not interested in the market for boxes, it might be thought that the above search could be extended to read: (market and (linerboard or "corrugating medium" or corrugated) not box. However, this would eliminate any records which refer even fleetingly to the box industry but which may still contain useful information.

The Paperbase database designates articles of less than one page by the phrase "Short article". The not operator may be used to eliminate these articles, if you are looking for something more in-depth, as follows: (market and (linerboard or "corrugating medium" or corrugated)) not "short article"

Note that the search is not case sensitive and that the inner brackets are resolved before the outer brackets.

Truncation can be used to expand a search from a word stem. In many databases, there are two types of stems: the question mark "?" used for a single character and the asterisk "*" used for an indefinite number of characters. Thus board? would retrieve board and boards, while coat* would retrieve coat, coats, coating, coatings, coater, coaters etc. The? may also be used to cover for both American and British English spellings such as alumin?um or colo?r.

Keywords, added by the database editor to indicate the principal subject-matter of the article, can be used for a more focussed search. Examples of typical keywords on Paperbase are cellulose, COD, kraft pulping and TCF. The full list can be obtained via the Paperbase Advanced Search Screen. A search can also be limited to a country or region, such as the United States or Europe or to records published in a range of years.

PUTTING THIS INTO PRACTICE:

* If you have too few hits, consider the use of plurals, synonyms and alternative spellings, or moving very specific keywords to the main search box

* With too many hits, think of restricting the search to appropriate years of publication or geographic locations, particularly for market or company news, adding keywords or removing short articles as above.

* Be prepared to have several tries at a search. Read through the results of your first attempt for clues for appropriate synonyms and keywords.

Finally, it is worth mentioning that Paperbase has live links to over 14,000 TAPPI articles, so you can instantly locate the full text of those which come up in your search.

Chris Torrero, Pira International

ABOUT THE AUTHOR

Chris Torrero is editor of Paperbase Abstracts and works for Pira International, Leatherhead, England. Contact him at [e-mail].

[ILLUSTRATION OMITTED]