PH 570: Introduction to Health Informatics

The Process

Information retrieval is often a continuous process during which you will consider, reconsider and refine your research problem, use various different information resources, information retrieval techniques and library services and evaluate the information you find. The figure below implies that the stages follow each other during the process, but in reality they are often active simultaneously and you usually will repeat some stages during the same information retrieval process.

The Plan

Before you can start planning your information retrieval, you need to define your research topic. Ask yourself what do you already know about the topic? What are the central concepts and theories relating to the topic? What question are you trying to answer?

Once you've defined your topic, you should consider where look for the information. Is it generic or scientific information that you're after? Are there specialist databases that you could use or can you find what you're looking for through Google? You should also start thinking about keywords and subject terms that you could use in your search. You'll achieve the best results in your information retrieval by using precise and specific search terms.

For example: 

The Techniques

Combining Search Terms with Boolean Operators (see above example)

AND 

• Use AND to combine your search words and narrow your search results
• all search terms must be present in the search results when you use AND
• For example if you want material that discusses both cats and dogs you would use the search cat AND dog. The red area in the middle of the Venn diagram below represents the result set for this search. It is a small set because the results have to include both search words.
• Remember that you can also combine phrase searches with Boolean operators, e.g. "" AND
• Also be aware that in many, but not all, databases and search engines the AND is implied. 

OR 

• Use OR to broaden your search results
• OR is commonly used to connect two or more similar concepts
• For example if you want material that discusses cats, dogs or both animals, you would use the search cat OR dog. All of the red area represents the result set for this search. It is a large set because any of the search words are valid using the OR operator.

NOT 

• Use NOT in a search to exclude search terms and narrow your search.
• For example if you only want to read about cats and nothing about dogs you would use the search cat NOT dog. The red area in the Venn diagram below represents the result set for this search. It is a small set using because NOT excludes all material with dogs from the result. 
• Be careful when using the NOT operator as it might limit the results too much and leave out some valid results.

Parenthetical Grouping (see above example)

If you are using more than two terms, then you must be aware of the logical order databases use to connect string of keywords and return results. Utilize parenthesis to connect multiple keywords and boolean operators together into a searchable string that the database understands. 

Phrase Searching Term

If you want to search for an exact phrase (search terms appearing next to each other) use quotation marks around the words, for example "wireless network".  If you don't use quotation marks around the words, many databases will look for the search words individually, so that the search results will include material where the search terms can be apart which may make the result irrelevant. The quotation marks also work in Google and Google Scholar searches. 

• For example: "heart attack" 

Truncation & Wildcards

Truncation, also called stemming, is a technique that broadens your search to include various word endings and spellings. Term truncation is particularly useful for languages with case endings, as it allows you to include all the inflected forms of the search term in the search. In LAB Primo the truncation symbol is the asterisk *.

• Example: using the search term librar* will include library, libraries, librarian in the search results. 

Wildcards substitute a symbol for one letter of a word. In LAB Primo the wildcard symbol is ? which is is used to replace one letter. Using wildcards is especially useful if a word is spelled in different ways, but still has the same meanings. 

•  Example: using the search term organi?ational will include both organizational and organisational in the search results.

MeSH and PubMed Specifics

Term Mapping

When users execute searches in PubMed, their keywords will automatically be mapped to equivalent MeSH terms.

• Example: A search for "cancer" in PubMed will return articles that contain the terms: cancer and neoplasms.

Automatic Explosion

PubMed uses a feature called 'automatic explosion' which automatically includes all of the MeSH terms located below your search term in the hierarchy. 

• Example: A PubMed search for "Eye" will capture the following terms: eye, eyelids, eyelashes, and eyebrows. 

Subheadings

Subheadings are topical qualifiers that are contained within every MeSH record. Subheadings, unlike MeSH terms, will not be automatically mapped. If you want to include them in your search, you will need to add them manually. Before adding a Subheading to the search builder, it is recommended to review the notes page to better understand how each Subheading is be defined in MeSH. Additionally, Subheadings will 'automatically explode' when added to your search. 

• For example: A search using the Subheading "epidemiology" will capture the following qualifiers: epidemiology, endemics, epidemics, frequency, incidence, morbidity, occurrence, outbreaks, prevalence, and surveillance