The Abstracts (AB) are always in English and use British spelling throughout. All important concepts are included along with any results. Authors' abstracts will often be used but will be amended, when necessary, in order to ensure that all the important points are covered. Some records will not include abstracts. There are two reasons for this. Some original articles, although significant, are short and require nothing more than a title to describe them while other items may be on the border line in terms of CABI's overall subject scope. While considered significant they may not warrant the extra effort required to prepare an abstract.

Searching the abstract field is ideal if you wish to retrieve absolutely every paper on a particular topic. This is particularly useful when searching for very new concepts where few records are anticipated. For the more common concepts, restricting the search to titles, descriptors and CABICODES is recommended. Here are a couple of search hints:

• Specific cultivars of crops are not generally allocated a descriptor, so searching the abstract for these in combination with the appropriate descriptor term is recommended. For example, to find important papers about the C57 cultivar of soybeans you should consider searching for soybeans in the Subject Heading field and combining it (AND) with C57 searched in the abstract:
  (soybeans.sh. AND C57.ab.)
• Concepts such as high temperature or specific techniques used in research, and so on, are best searched in the abstract.
• For maximum retrieval search the abstract bearing in mind it will, by definition, produce more false drops (low relevance).

The Additional Authors field (AD) is used for the input of the names of the individual authors of parts of documents, such as books that are being treated as a single record/item. This field is searched within the Author (AU) field.

The Addtional Title Data (AT) field contains data that is not appropriate for any of the other title fields, for example series statements for books and the titles of special issues of a journal and details of alternative series in which the record/item is also published.

This field is word indexed in Title (TI) and, when available, displays in Title (TI)

The Author (AU) field is used for the input of the names of all the individual authors of the item being cited. In this field the names are entered in the preferred format set by the editorial rules. However, the user of the database cannot be expected to know or remember all these complex rules, so the Author Variant field is used for awkward names where the availability of variations would be useful for searching. Both these fields are combined for searching purposes so, as long as one of the options provided is selected, the record will be retrieved.

Editor names, which are input in two separate fields, follow the same editorial rules as authors names and are combined with the author names in the Added Entry Personal Name (AE) alias. To search for a person who has authored and/or edited items in the database, search that person's name with an .ae. search qualification.

The basic format for the input of author and editor names is:
surnamefirst initialsecond initial

Titles of honour such as Sir are transposed but academic titles like Dr and Professor are omitted. Relationship terms like Sr., Jr., II, III, IV, etc. are entered after the initials except for Portuguese names (Neto, Junior, Filho, Sobrinho, etc.) where they are cited in full immediately after the surname.

All separate particles such as de, Di, Le, von, van, etc. are transposed to follow the forename initials. Arabic names are standardized by inserting hyphens after the particles Al- and El-, so that they become part of the surname. Here are some examples:

• Hoch, H.
• Hagedorn, DJ
• D'Arcy, A.
• Vleck, VE
• van Rooy, J.
• de El-Ragae, AH

When searching for author or editor names you should familiarize yourself with the way they are indexed on the system you are using. It is also worth noting that names are phrase parsed so it is advisable to use truncation or the index expand/look-up feature. Searching simply for a surname without allowing for all the initials will usually result in few or no records being retrieved.

Up-posted descriptors are the broader thesaurus terms for specific groups of the descriptors which have been assigned to a record at its particular level of specificity. The selection of these terms is driven by hierarchical relationships present in the CAB thesaurus. They are applied only to the descriptors that appear in the Organism Descriptors (OD) field and the Geographic Location Descriptors (GL) field.

The up-posted descriptors help in finding items which deal with any instance of a broader concept. This is not the same as searching for items which deal broadly with that concept.

For example, the term Africa as a Geographic Location Descriptor should retrieve all items which deal with the continent as a whole. A search for the same term among up-posted descriptors will retrieve records specific to Ghana or Algeria or East Africa or any other individual country or region in the continent.

The up-posted descriptors are held in a different field from the descriptors applied at a document's own level of specificity, to avoid cluttering searches for broad-level concepts with unwanted items indexed at the more specific level. For example, a searcher looking for general texts on bacteria can search for this term as a descriptor term (for example, bacteria.de.) and avoid retrieving thousands of records on specific bacteria. In other words, the searcher is given the option of going for higher recall without losing the option of good precision.

Up-posting is an extremely complex operation. It cannot be automatically applied to every term in every hierarchy in the thesaurus because many terms have more than one meaning with the consequent, automatic, generation of a great many erroneous entries. Neither can up-posting be applied manually for all terms. The vast amount of extra editorial time and keyboarding would severely reduce the timeliness of the database. It is for this reason that the system of automatic up-posting has only been applied to organism names and geographic names.

Organism names (entries in the OD field)
All Organism Descriptors are up-posted using the full taxonomic hierarchies from the new CAB Thesaurus. Hierarchies up-post to:

Geographic names (entries in the GL field)
All terms in the Geographic Descriptor field are up-posted using the geographic and geopolitical hierarchies from the new thesaurus. Hierarchies up-post to one or more of ACP Countries:

Corporate authors, such as the World Health Organization, are entered in a field of their own. They are made searchable separately from personal authors/editors. It is impossible to impose any strict editorial control over the data entered into this field so, when searching, it is important to take into account any abbreviations and acronyms that may be used as well as any known name changes that may have occurred within the time limits of the particular search in question. In an attempt to alleviate the problem of inconsistency there is also a Corporate Author Variants field into which variations may be put. These two fields are merged for searching purposes.

CABICODES are classification codes applied to all records to indicate the broad subject areas within which they fall. Altogether there are about 270 CABICODES, of which any one will typically retrieve thousands of records. They consist of two letters followed by three numbers. They are particularly useful for isolating a subset of the database within which a more specific search may be performed. They are ideally suited for the retrieval of records in areas that are difficult to define using keywords. For example, a search on VV300 will retrieve all papers on public health and nuisance pests from the point of view of human health and hygiene. CABICODES complement keyword indexing. This means to say that a concept which is allocated a CABICODE will also be indexed with all the appropriate keywords from the thesaurus.

Every record will have at least one CABICODE representing the main subject of the paper. In addition, relevant codes will be added for other important concepts that are explicitly discussed in the paper. Subjects reported by implication only will not be coded. For example, a paper on nitrogen nutrition of mulberry would be coded FF061 (PLANT NUTRITION) but not LL020 (SERICULTURE) unless the effect of the nitrogen nutrition on silkworms was actually discussed.

The codes are arranged in 20 distinct groups. Each group has a general code with its corresponding classification name. This is then followed by more specific codes and names. As with the keyword indexing the codes are allocated at the most appropriate level of specificity. CABICODES can be searched in full or with truncation.

This field is also browsable by way of a CABICODE classification code Tool, and can be searched by word in the CABICODE Word (CW) index.

Codes are cascaded, or exploded in the Exploded CABICODES (XC) field. Searching on VV000 in the CC field will only retrieve general papers on Human Health and Hygiene. It will not pick up all papers with a VV code. To do this you could search on VV truncated (for example, VV$.cc.), or, more simply and quickly search the exploded form (for example, VV000.xc.)

CABICODES are classification codes applied to all records to indicate the broad subject areas within which they fall. Altogether there are about 270 CABICODES, of which any one will typically retrieve thousands of records. They consist of two letters followed by three numbers. They are particularly useful for isolating a subset of the database within which a more specific search may be performed. They are ideally suited for the retrieval of records in areas that are difficult to define using keywords. For example, a search on VV300 will retrieve all papers on public health and nuisance pests from the point of view of human health and hygiene. CABICODES complement keyword indexing. This means to say that a concept which is allocated a CABICODE will also be indexed with all the appropriate keywords from the thesaurus.

Every record will have at least one CABICODE representing the main subject of the paper. In addition, relevant codes will be added for other important concepts that are explicitly discussed in the paper. Subjects reported by implication only will not be coded. For example, a paper on nitrogen nutrition of mulberry would be coded FF061 (PLANT NUTRITION) but not LL020 (SERICULTURE) unless the effect of the nitrogen nutrition on silkworms was actually discussed.

The codes are arranged in 20 distinct groups. Each group has a general code with its corresponding classification name. This is then followed by more specific codes and names. As with the keyword indexing the codes are allocated at the most appropriate level of specificity. Ovid allows for the codes to be searched in full or with truncation.

This field is browsable by way of a CABICODE Classification Code Tool, and can also be searched in the CABICODE (CC) and Exploded CABICODES (XC) fields.

Codes are cascaded, or exploded in the Exploded CABICODES (XC) field. Searching on VV000 in the CC field will only retrieve general papers on Human Health Hygiene. It will not pick up all papers with a VV code. To do this you could search on VV truncated (for example, VV$.cc.), or, more simply and quickly search the exploded form (for example, VV000.xc.)

The Descriptor Index (DE) contains the Geographic Locations (GL), the Organism Descriptors (OD) and Subject Headings (SH).

The Document Title (DT) field is the title of the complete document in which the abstracted item was published. If the item is an individual paper from a conference proceedings then the document title will be the title of the proceedings.

The names in the Editor (ED) field are entered in the preferred format set by the editorial rules. However, the user of the database cannot be expected to know or remember all these complex rules, so an Editor Variant field is used for awkward names where the availability of variations would be useful for searching. Both these fields are combined for searching so, as long as one of the options provided is selected, the record will be retrieved.

Editor names follow the same editorial rules as authors names and are combined with the author names in the Added Entry Personal Name (AE) alias. To search for a person who has authored and/or edited items in the database search that persons name with an .ae. search qualification.

The basic format for the input of author and editor names is: surname, first initial second initial

Titles of honour such as Sir are transposed but academic titles like Dr and Professor are omitted. Relationship terms like Sr., Jr., II, III, IV, etc. are entered after the initials except for Portuguese names (Neto, Junior, Filho, Sobrinho, etc.) where they are cited in full immediately after the surname.

All separate particles such as de, Di, Le, von, van, etc. are transposed to follow the forename initials. Arabic names are standardized by inserting hyphens after the particles Al- and El-, so that they become part of the surname. Here are some examples:

• Hoch, H.
• Hagedorn, DJ
• D'Arcy, A.
• Vleck, VE.
• van Rooy, J.

When searching for author or editor names you should familiarize yourself with the way they are indexed on the system you are using. It is also worth noting that names are phrase parsed so it is advisable to use truncation or the index expand/look-up feature. Searching simply for a surname without allowing for all the initials will usually result in either very few or no records being retrieved.

The Every Title (ET) alias searches the (English) Title (TI) and the Original Title (OT) fields.

Titles of original items fall into one of two categories: English and non-English. The non-English titles are further divided into Roman and non-Roman script. Global Health is an English language database and, for this reason, all records in it are given an English title regardless of the language in which the original document was written. Where possible a non-English title is also provided to aid retrieval. There are two separate fields provided for data input; the (English) Title field (TI) and the Original Title field (OT).

The Geographic Location (GL) descriptor field is controlled by the CAB Thesaurus.

Geographic Location descriptors describe the countries and subdivisions, such as states or provinces, that are relevant to the record. The field contains only controlled terms from the latest version of the CAB Thesaurus and these are manually allocated to the record at the stage of record creation. The object of applying vocabulary control is to label each important concept in the source item with one, unambiguous name, preferably the name a searcher would choose for the same concept. The thesaurus lists all the allowed descriptors and provides guidance as to the scope of each. Precision and recall are both enhanced if the thesaurus is consulted by searchers as well as indexers.

All the terms in the Geographic Location field are automatically upposted so that all their higher thesaurus terms are added to the separately searchable Broad Term (BT) field.

To search Geographic Location by word, use the Heading Word (HW) index.

Sometimes you may wish to retrieve every Subject Heading (SH), Organism Descriptor (OD), Broad Term (BT), Geographic Location (GL) and Identifier (ID) that includes a particular word or words; this is done by searching the desired word(s) in the Heading Words (HW) field.

The ISBN (IB) field contains the International Standard Book Number (ISBN). Spaces and dashes have been removed from the display and index to facilitate linking capabilities.

The field contains free language words or phrases assigned by indexers to give fuller notation of the document content than are provided by the controlled index terms. Often, the Identifer field contains new terms, for example the names of new species, which are not yet in the CAB Thesaurus.

Identifiers are word indexed in Heading Word (HW).

The Institution (IN) or Author Affiliation field is used for the addresses of both personal and corporate authors. Details entered in this field include institution name, town, and country when available. Where possible, the full address is now entered and will include post and zip codes. The field itself is intended to give the address of where the work was done. For personal authors the address is that of the first author. If a correspondence address is given with a paper it is only cited if no author address is present, and when used the phrase "correspondence address" is appended in the record display.

As with the corporate author field, it is not possible to apply rigid editorial rules to this field so, when searching, it is important to take into account all possible variations and combinations, including name changes and abbreviations. Some examples are:

• Cornell University, Ithaca, NY 14853, USA
• IRCT, 75116, Paris, France
• Operations Research Division, EMBRAPA, 70-333 DF, Brazil
• ICARDA, Aleppo, Syria

The ISSN (IS) field contains the International Standard Serial Number for the journal in which an article was published.

The ISSN is the International Standard Serial Number and is a unique number identifying serial publications such as journals. The ISSN consists of two groups of four digits in Arabic numerals, except possibly for the last, check digit, which may be an X. This conformity makes it easy to search for ISSNs, which can be entered with or without the hyphen between the two groups of four characters.

It appears as an 8 digit number, separated by a hyphen: 0028-4793.

The Journal Name (JN) field contains the full name of the journal in which an article was published.

Journal names are indexed as phrases, so enter enough letters of the journal name to locate the name in the index: new engl$ (for New England Journal of Medicine).

To search by word, use the Journal Name Word (JW) Index.

The Journal Name Word (JW) contains each word in every journal name indexed in Global Health. Enter the single most descriptive word in a journal title.

A single heavily posted word, such as "journal" will search slowly.

The Language (LG) field contains the original language(s) of the entire document. Search for the full name of the language. For example, to find documents originally published in French, search for French in the LG field.

The Local Messages (LM) field contains messages created by the System Administrator to indicate information about journals held in your local library system.

The Location of Publisher (LP) field includes the geographic location of the publisher. It displays as part of the Source (SO) field.

The Meeting (MT) field will contain any or all of the following data if available: the title of the conference, its location and dates. There is no strict editorial control regarding the way the data is entered so all possible variations should be accounted for when searching. Enter the single most descriptive word in a conference name, location or date. Consider both full spellings and abbreviations.

The Notes (NT) field contains any additional, supplementary or explanatory information about the document.

The field is meant for display purposes only, although it is searchable.

The Organism Descriptors (OD) field is controlled by the CAB Thesaurus.

The majority of the names are taxonomic; common names are used for livestock and common crops, and some well known wild organisms. The field contains only controlled terms from the latest version of the CAB Thesaurus and these are manually allocated to the record at the stage of record creation. The object of applying vocabulary control is to label each important concept in the source item with one, unambiguous name, preferably the name a searcher would choose for the same concept. The thesaurus lists all the allowed descriptors and provides guidance as to the scope of each. Precision and recall are both enhanced if the thesaurus is consulted by searchers as well as indexers.

All the terms in the Organism Descriptors field are automatically upposted so that all their higher thesaurus terms are added to the separately searchable Broad Term (BT) field.

The Original Title (OT) field contains the item's original title when the item was published in a language other than English.

Titles of original items fall into one of two categories: English and non-English. The non-English titles are further divided into Roman and non-Roman script. Global Health is an English language database and, for this reason, all records in it are given an English title regardless of the language in which the original document was written. Where possible a non-English title is also provided to aid retrieval. There are two separate fields provided for data input; the (English) Title field (TI) and the Original Title field (OT).

To search both the (TI) and (OT) fields simultaneously, use the search alias Every Title (ET), such as academy.et..

For non-English items with non-English titles the OT field contains the original language title provided that it is in Roman script (for example, French, German, Italian, and so on). It is not possible to provide original language titles for item titles in non-Roman script (for example, Greek and Cyrillic) and the current editorial policy states that no transliterated versions will be provided. However, users will find a few transliterated titles in older records on the database.

This field contains any price information which was given with the original item and displays as part of the Source (SO) field.

All records in the database are designated at least one of the following publication type categories:

The field is mandatory and is validated against an authority file. The field has also been added to the back file records and entries created automatically. However, as with any automatic system, there were inevitably some records which could not be recognized so these are all grouped in with the "Miscellaneous" records.

The Publisher (PU) field contains the name of the publisher. This field usually displays as a part of the Source (SO) field. The information is not standardized. Consider both full spellings and abbreviations.

The CAS Registry Number (RN) field contains the Chemical Abstracts Registry Number for all chemical names entered in the Subject Heading (HW) field for which a corresponding registry number exists.

Entries are generated automatically from an authority file. Through the backfile conversion, entries have been generated for all the backfile records.

The Subject Headings (SH) field is phrase indexed and controlled by the CAB Thesaurus.

The field contains only controlled terms from the latest version of the CAB Thesaurus which describe the key points of the paper and these are manually allocated to the record at the stage of record creation. The object of applying vocabulary control is to label each important concept in the source item with one, unambiguous name, preferably the name a searcher would choose for the same concept. The thesaurus lists all the allowed descriptors and provides guidance as to the scope of each. Precision and recall are both enhanced if the thesaurus is consulted by searchers as well as indexers.

The SH field contains any Thesaurus term that is not considered to be an Organism Descriptor or a Geographic Location descriptor. Examples would be terms like Pesticides, Nutrition, Leisure Management, etc. Refer to the notes on Organism Descriptors (OD) and Geographic Location (GL) descriptors for details of what these two, separate index fields contain.

To search Subject Headings by word, use the Heading Word (HW) index.

The Secondary Journal Source (SJ) field is phrase indexed. If CAB International did not see the original publication, the secondary source is cited here. The SJ field usually displays as part of the Source (SO) field.

The Source (SO) field includes a display of all the basic information needed to locate a citation, including the Journal Name (JN) or Document Title (DO), Publisher (PU), Location of Publisher (LP), Country of Publication (CP), Date of Publication (DP), Volume (VO), Bulletin/Issue/Part Number or Edition field (IP), Page Numbers (PG), Price (PR), References (RF), and Secondary Journal Source (SJ).

The Title (TI) field contains the item's English-language title.

Titles of original items fall into one of two categories: English and non-English. The non-English titles are further divided into Roman and non-Roman script. Global Health is an English language database and, for this reason, all records in it are given an English title regardless of the language in which the original document was written. Where possible a non-English title is also provided to aid retrieval. There are two separate fields provided for data input; the (English) Title field (TI) and the Original Title field (OT).

To search both the (TI) and (OT) fields simultaneously, use the search alias Every Title (ET), such as academy.et..

If the original document is in English with an English title then this title is input exactly as given. Subtitles are included and any abbreviations or American spellings are not altered. The only changes that are allowed are corrections of obvious spelling errors which are important for retrieval purposes. In a small percentage of cases some titles may be annotated to improve sense or clarity.

If the original document is not in English but carried an English language version of the title this will be used, as far as possible, exactly as given. It will be edited only as necessary to clarify either sense or grammar or to correct any obvious spelling errors.

If no English version is available, or if the English version provided diverges a lot from the non-English version, the latter is translated, by CABI, into English.

The Year of Publication (YR) field contains the year in which a document was published. The year may be searched using four digits, such as 1994.yr. Some records may contain more than one year because of the special circumstances with the document's publication date.

Each year is indexed separately. For more information about the date of publication, please consult the Source (SO) field.