Subject Headings Search Help
   

Contents

INTRODUCTION
The Classification Web Subject Headings software is used to search., display and navigate the Library of Congress Subject Heading (LCSH) data using a standard Internet Web browser.  It provides hypertext links to related subject areas, as well as other electronic resources.  This includes links to the classification browser, subject correlations screen and to selected bibliographic records.

Basic system requirements for your computer and a description of the classification components of this site are located in the "Getting Started with Classification" section of the help system.

The Subject Search page is used to both search and navigate subject headings.   Once subject headings are displayed on the screen, you can navigate between subjects and move from page to page using the links and buttons provided.  Many of the links in Classification Web open new web browser windows so that you can explore various avenues without loosing your place.  To prevent your computer screen from getting overwhelmed, remember to close any Classification Web windows when you are finished with them.

OPTIONS
Certain options that you select while using Classification Web are passed from screen to screen by the software.  This has two implications:  First, different windows can be open at the same time that were started with different options.  Second, these windows pass their current options to any new windows that they open.  These options include the number of records to return with each request and the type of classification browser you prefer.

You can choose default settings for these options by clicking on the preferences link of the main menu.  You may also find occasion to use the extended menu.  This menu allows you to select a different style of classification browser (standard, enhanced or hierarchy) without changing the default selection  for your account.  The classification browser types are described in more detail here.

STRUCTURED vs. UNSTRUCTURED HEADINGS
The Subject Heading database supports two types of access points for displaying subject data: structured and unstructured.  The structured approach uses similar rules to the existing subject heading print product.  That is, all of a subject's subdivisions sort before moving on to the next subject.  The unstructured index interfiles subjects and their subdivisions with other subjects.  In other words, when you are using a structured subject heading index, you have to know how a subject was subdivided in order to find it easily.  Whereas, with the unstructured index, you don't.  One result of this distinction is that you must input double dashes (--) between subject headings and their subdivisions when using a "structured" subject index.

Take for example the subject term of "railroads".   If you use the "structured" subject heading index, you have to navigate past all of the subdivisions of "railroads" until you get to "Railroads, Aerial".  The sequence is something like this:

Railroads--Zaire
Railroads--Zambia [R S]
Railroads--Zimbabwe [R S]
Railroads, Aerial
Railroads and earthquakes
Railroads and state (May Subd Geog) [R S]
By using the "Unstructured subject heading" index,  the term "Railroads, Aerial" follows "Railroads--Advertising".  The sequence looks like this:
Railroads--Advertising
Railroads, Aerial
Railroads--Africa, East
Railroads--Africa, Northeast
Railroads--Air-brakes
Railroads--Airspace utilization
Railroads and earthquakes
Railroads and state
In other words, the "unstructured" index sorts in straight dictionary order, whereas the "structured" index is hierarchical.  The "unstructured" index is easier to search because you don't have to input dashes in all of the right places.  However, the data that is returned will not be in the order that has been traditionally used in the LCSH print product.

These two types of access points are also provided for free-floating subdivisions.

QUERIES vs. BROWSING
The Subject Search screen supports two ways to explore subject headings: by running queries and through index browsing.  A query allows you to search for one or more criteria using boolean logic (AND, OR and NOT), mathematical operators (greater than, less than, etc.) and wild card characters (to match multiple characters).  The result is a set of records that meet the search criteria, returned in no particular order.  To run a query, input your criteria into one or more of the input fields on the search screen and then press the Search button.

Microsoft Internet Explorer Users
If you are using Microsoft Internet Explorer (MSIE) as your web browser, be careful about pressing the ENTER key on any of the Classification Web search screens.  In general, pressing ENTER in MSIE on a form causes the form to be submitted to the server.  In this case, Classification Web will think you clicked on the Search button.  You will not get the same results as if you click on the appropriate Browse button.
Index browsing, on the other hand, allows you to choose an existing database index (like subject heading , keyword, classification number, etc.), input a starting value and jump to that place in the index.  The results are returned in "index order" (usually this is alphabetical order) so that you can navigate forwards and backwards in the index to locate the exact term you are looking for.  Index browsing does not support boolean or mathematical operators, or wild card characters.

Index browsing is almost always faster and easier than running queries, so it is recommended that you start by browsing and move onto queries as necessary.  You browse an index by inputting your search term into the appropriate field on the search form (like caption or keyword) and pressing the Browse button immediately to the right of the field.  There is a Browse button corresponding to each input field on the search screen.  Make sure that you click on the appropriate Browse button, otherwise you will get back the wrong information.

Terminology
Whenever this help document refers to a "search" as in "search screen", "running a search" or "search results", it includes both index browsing and actual queries.  If any part of this document refers to only index browsing or only queries, that will be explicitly stated.
Response Time
Index browsing will always produce an almost immediate response back to your computer.  This is because the server has to do very little work in returning the contents of an index.  Running a query, however, may take noticeably longer.  The length of time it takes for query to run depends on the number of records that match each one of your criteria.  Even if the final result is only a handful of records, if any one of your search terms would have resulted in tens of thousands of matches on its own, you may experience some delay.

If you are using Netscape Navigator as your Web browser, the server will return a status screen every ten seconds (if the query takes that long) to show you what its progress is.  Unfortunately, Microsoft Internet Explorer does not support the protocol that makes this possible.  Instead, the server will return a space character every ten seconds.  This is to stop the browser from thinking that the connection to the server has gone down (your screen will remain blank, however).

It is important when you are running a query that you don't press the ESCAPE key, click on your Web browser's Stop button or use any other button (backwards, forwards, etc.) that would cause the browser to replace the contents of the query screen.  If you do, you will cancel the query in progress because all of these actions will cause your Web browser to drop the connection to the server.  Whenever the connection between your Web browser and the server is interrupted, the server will stop whatever it is in the middle of, reset itself and wait for your next request.  If a query is taking a while and you want to do something else, just use another web browser window or work with an entirely different piece of software (word processing or e-mail).

When you do run a query, keep in mind that the more specific your search terms are, the faster the query will run.  The fewer records that match each search criteria the faster the query runs.  This means you should take particular care in your use of wild card characters, particularly when placed near the front of a search term.  In fact, the use of leading truncation (placing a * or ? at the front of  a search term) will generally result in significantly slower queries.

Punctuation
Whether you are index browsing or running a query, this server treats almost all punctuation (comma, period, hyphen, dollar sign, parentheses, slash, etc.) as a space and multiple spaces as a single space.  In addition, any leading and trailing spaces (or  punctuation) are ignored. Any punctuation that is not treated as a space is ignored (apostrophe and caret: ' and ^).

This makes it possible to find what you are looking without knowing how the original record was punctuated or formatted.  In classification number indexes the following punctuation is not ignored: colon, hyphen and period.

INDEXES
Each of the input fields on the search form correspond to a different database index or access point.  The indexes are:

Structured Subject Heading
This index includes all subject headings and their tracings (links to other subjects), sorted in traditional LCSH order: a subject's subdivisions come before the next subject is printed.  See Structured vs. Unstructured, above for more detail.

This index is sorted using the following fields:

        100, 110, 111, 130, 150, 151, 155,
        400, 410, 410, 430, 450, 451, 455,
        500, 510, 510, 530, 550, 551 and 555

From each of these fields the following subfields are extracted for the subject heading:

        a-h and j-u

And the following subfields are extracted for the subdivision:

         v, x, y and z

Unstructured Subject Heading
This index includes all subject headings and their tracings (links to other subjects), sorted in dictionary order: a subject's subdivisions may be interfiled with other subjects, unlike the LCSH print product.  See Structured vs. Unstructured, above for more detail.

This index is sorted using the following fields:

        100, 110, 111, 130, 150, 151, 155,
        400, 410, 410, 430, 450, 451, 455,
        500, 510, 510, 530, 550, 551 and 555.

From each of these fields the following subfields are extracted for the subject heading:

        a-h and j-u

And the following subfields are extracted for the subdivision:

         v, x, y and z

Structured Free-floating Subdivision
This index includes all free-floating subdivision fields and their tracings (links to other subdivisions), sorted in traditional LCSH order: a subdivision's subdivisions come before the next subdivision is printed.  See Structured vs. Unstructured, above for more detail.

This index is sorted using the following fields:

        180, 181, 182, 185,
        480, 481, 482, 485,
        580, 581, 582 and 585.

Along with the following subfields:

        v, x, y and z

Unstructured Free-floating Subdivision
This index includes all free-floating subdivision fields and their tracings (links to other subdivisions), sorted in dictionary order: a subdivision's subdivisions may be interfiled with other subdivisions, unlike the LCSH print product.  See Structured vs. Unstructured, above for more detail.

This index is sorted using the following fields:

        180, 181, 182, 185,
        480, 481, 482, 485,
        580, 581, 582 and 585.

Along with the following subfields:

        v, x, y and z

Keyword
Only those fields that are included in the "subject heading" and "free-floating subdivision" have been included in the keyword index.  This system does not use "stop lists" in generating the keyword index. Therefore, words like "and", "or", "a", "an" and "the" are all included in this index.

Any queries that you run that involve the keyword index are handled differently than any of the other indexes.  Because the keyword index only contains individual words, if you input more than one word in the keyword field and run a query (by clicking on the Search button), the software will search for those words separately in the same record.  There is no support for phrase searching in the keyword index.  Putting in more than one word and clicking on the Browse button will just take you to the first word in your list.

The keyword index is the only index that you don't have to use boolean mode to search for multiple values at the same time.  If you input more than one word (words are anything separated by a space), the program treats each word as a separate criteria.  Whether or not a record matches your criteria depends not only on the words you input but also on whether you have selected the AND or the OR search option.  If you have selected the AND search option, every word in the keyword list must be found in the record for the record to be a match.  When you select an OR search, if any of your words can be found, then that record will match.

The following fields are included in the keyword index:

        100, 110, 111, 130, 150, 151, 155, 180, 181, 182, 185,
        400, 410, 410, 430, 450, 451, 455, 480, 481, 482, 485,
        500, 510, 510, 530, 550, 551, 555, 580, 581, 582 and 585.

Along with the following subfields:

        a-h, j-u, v, x, y and z

Classification Number
This index contains the classification number from each record that has such a number.  Not all subject headings have a classification number assigned to them.

The classification number index contains the following fields:

        053/a and 053/b

Record Number
This index contains the 001 field -- control number -- from each LC subject heading record.  This is the same value as the 010 field -- Library of Congress Control Number-- in these records.

THE DISPLAY
The search screen is made up of five to seven sections depending on whether there are any data records on the screen.  When you are browsing an index or looking at the results of a query, the data is displayed at the top of the page and the rest of the search screen follows.  The sections are, from top to bottom:

Navigation and Command Buttons
The navigation buttons that display other pages of search results (first, next, previous and last page commands) are only displayed after you run a search that returns some records.  When you first reach the search screen, these navigation buttons will not be shown.  The possible navigation and command buttons are:
 
Button Action
|< Returns the first page of search results.
< Returns the previous page of search results.
> Returns the next  page of search results.
>| Returns the last  page of search results.
Search Runs a query using the values input on the screen.
Reset Clears any values from the input fields on the search screen and resets all display and search options to their default values.
Logout Logs you out of Classification Web.
Close Closes the current window.  In cases where the software has opened a new window for you (like when you click on the Search button), you can close the window by clicking on this button.  If the software is not able to close the window because it was opened by hand, the main classification menu is displayed instead.
Menu Displays the main menu.
Help Displays on-line help in a separate window.

Search Results
The first line after the command buttons gives the name of the search screen you are using (Basic or Advanced) and information about the index or query you are using, along with the number of items found  in the case of a query.  Here is what might be displayed when browsing the caption index:

   Subject Search: Unstructured subject heading

The number in parentheses are the total number of captions in this index.  Remember, this number will vary widely, depending on whether you are working with the whole database or individual subsets.

A query might produce the following:

    Subject Search: query1 (3 items)

This query located three records that matched the search criteria.

Following the search information are the search results.  If you are browsing one of the subject heading or free-floating subdivision indexes, the screen will look most like a regular subject heading list.  This is because tracings (links to other subjects) will be interfiled in the appropriate places.  If you choose one of the other indexes or run a query, the records will be displayed out of their usual context.

Subject Heading Abbreviations and Links
Subject records contain links to other subject headings in the database.  The following abbreviations are used:
 

Abbreviation Meaning
BT Broader term.
NA Not applicable.
NT Narrower term.
SA See also.
UF Use for.
USE Use.

Any classification numbers that are displayed on the subject screen are links to the classification browser for that number.

Additional Links
Each subject heading record that is displayed includes a number of links.  The classification number for each record is a link to the classification browser for that particular number.  When you click on a number, the classification browser will display the  classification schedules starting with that number.

At the end of many of the subject heading lines on the screen are a set of letters in square brackets.  For example, here is the entry for "arctic conditions":

Arctic conditions [R B S L]
These links are used to perform actions related to this particular record.  The possible letters are:
 
Button Action
R Displays the underlying subject authority record in a separate window.
B Searches for and displays any bibliographic records with this classification number in a separate window.
S Runs a subject correlation search in a separate window for this classification number.
L Index browses for any subject records that use this classification number. 

Search Input Fields
The input fields are used to enter your searching criteria.  If you are only using one field and you don't need wild cards, boolean logic or mathematical operators, click on the Browse button to the right of the field to start an index browse.   Otherwise, fill in your various criteria and click on the Search button.

Each of the input fields correspond to a different classification indexes, which are described above.  Documentation on wild cards, boolean logic and mathematical operators can be found below.

Display Options
You can set default values for all of the following display options by clicking on the preferences link of the main menu:

Classification Browser Options
If you click on a link that launches the classification browser, the browser display style that is used is determined by your user preferences and the menu that you used to get to the current screen.  You choose your default settings for the system by clicking on the preferences link of the main menu.  You can override your default browser type by going through the extended menu.  The classification browser types are described in more detail here.

Search Options
The search options only affect queries and not index browsing.  The options are:

AND vs. OR Search
This option only affects your searches if you have filled in more than one field with criteria and if you are searching the keyword index.  In these cases, you can ask for those records that meet every criteria you entered (this is an AND query) or you can ask for those records that meet any of the criteria you input (an OR query).

This is an example of an AND query:

        Find all records that have "products" in the keyword index AND a classification number that starts with "HF".

To run this query, you would make sure the AND option is selected (it is by default) and fill in the input fields as follows:

 
Input Field Input Value
Keyword products
Classification number HF
This query will return around 16 "hits".

Now try the same query with an OR operator.  To run this query, you would make sure the OR option has been selected before you click on the Search button.  It should result in approximately 1366 "hits".

Simple vs. Boolean Search
So far, this document has only described how to to input a simple query.  This is the default setting for the search screen.  However, your searches can become much more powerful if you turn on the boolean option.  In boolean mode, you can have more than one search term per input field , and you can use wild card characters and mathematical operators.

The most important  difference between inputting a boolean query and a simple query is that you have to put quotation marks around each of your search terms when using boolean mode.

Here is an example of running a keyword query for multiple terms (the Boolean option must be selected for this to work):

 
Input Field Input Value
Structured Subject Heading "statistics" or "finance"
Keyword "administration" or "insurance"
The outcome of this query will also depend on whether you have chosen the AND or OR search option described in the previous section.  The AND and OR option controls the relationship between these fields.  If AND is selected, a record would have to have either of these two captions AND either of these two keywords.  If OR is selected, a record could either have one of these captions OR one of these keywords in order to be a match.

Here is an example using boolean logic and mathematical operations:

 
Input Field Input Value
Classification number (>= "KJ" and < "KJU") or ( >= "KKE" and < "KL")
This is the query that is used for the KJ-KKZ  subset (European Law).  It illustrates how parentheses can be used to insure the proper grouping of search terms and the use of mathematical operators.  Note that if mathematical operators are used, they must be placed immediately before the search term.  If you don't use a mathematical operator, EQUALS (=) is assumed.

Each query operator has a precedence which determines how the software interprets your query.  This is no different than the use of precedence in algebraic formulas.  For example:

        4 + 5 x 6

The multiplication operator has a higher presence than the plus operator.  As a result, you multiple five and six before adding the four.  In the preceding example, it turns out the parentheses are not needed.  This is because AND has a higher precedence than OR.  Parentheses don't hurt, however, and they often make a query more readable.

Boolean Operators
Here is the list of query operators and their relative precedence values:

 
Operator Precedence Meaning
() 7 Parentheses
not 6 Logical NOT
cmp 5 Compare operator
< 4 Less than
> 4 Greater than
<= 4 Less than or equal to
>= 4 Greater than or equal to
= 3 Equal to
!= 3 Not equal to
<> 3 Not equal to
>< 3 Not equal to
and 2 Logical AND
or 1 Logical OR
The compare operator (CMP) is only included for completeness.  It is used by the software to turn the information on the Basic Search screen into a query that this server can use.  You may have noticed that in all of the query examples, you never had to input a field name -- it  was done for you at some stage.  When your query is submitted to this server, the appropriate field name is inserted along with the CMP operator and your search value.  Only if you use the Advanced search screen to submit a free-form query will you ever need to use the CMP operator.

Case Sensitive Search
Case sensitivity refers to how a computer handles upper and lower case letters when searching for data.  With this server, it also includes the issue of diacritical marks like accents and umlauts that are used in non-English languages.  By default, this server uses a relaxed approach with search criteria when dealing with upper/lower case issues and diacritical marks.   As a result, you don't have to worry about how the data is capitalized and accented when you input your search terms.  Please note that any time you use a Browse button to browse an index, the lookup is always relaxed.  You cannot perform a case sensitive index browse -- you must run a query instead.

If you are looking for a specific capitalization and accenting, there are two procedures to follow depending on whether you are in Boolean mode or not.  If you are not in Boolean mode (i.e. the "Simple" search option is selected), you can check the "Case sensitive search" button before pressing the "Search" button.  If you are in Boolean mode, you must use back-quote characters around your search terms instead of double-quotes.  The back-quote character is on the same key as the tilde (~).  Here is an example:

 
Input Field Input Value
Structured Subject Heading "Statistics" or "Finance"
Keyword "Administration" or "Insurance"
As you can see from this example, each search term can have a different case sensitivity requirement when using Boolean mode.

Truncation Search
By default, every search term you input in a query is treated as if it was just the prefix or beginning of the word or phrase you want.  This is sometimes called a truncation search.  So if you are searching for "statistic", the system will also match on "statistics", "statistical", "statistician", etc.  This makes it easier to find all variations of a word without having to enter all of them in.

If you are looking for a specific version of a word or phrase, you can select the "Exact match" option at the bottom of the search screen.

Wild Card Characters
The truncation search that is described in the previous section is an example of the use of wild card characters.  this server supports two such characters: the asterisk (*) and the question mark (?).  An asterisk matches zero or more characters and a question mark matches exactly one character.  You can accomplish the same thing as a truncation search by selecting the "Exact match" option and adding asterisks to the end of each of your search terms.

You can use wild card characters in both the simple and boolean modes.   Here is an example with a simple search:

 
Input Field Input Value
Unstructured Subject Heading rail* employees
Here is the same example in boolean mode:
 
Input Field Input Value
Unstructured Subject Heading "rail* employees"
Both of these queries match any caption that starts with "rail" and ends with " employees".  In the classification database, both "railroad employees" will be found.

Things to remember:

Additional Command Buttons
The additional command buttons at the bottom of the basic search screen include:
 
Button Action
Redisplay Redisplays the current screen.  Use this button after you have used the display option that controls how much of each record's hierarchy is displayed.

Return to the top of this page.

Copyright © 2001 Minaret Corp. All rights reserved.