- Introduction to the Simple Search
- Getting to know the Search screen
- Quick Search (search and go)
- Complex search using the simple search screen
- Formulating a search
- Introduction to the advanced search
- Map-based Search
- STOP words (words that are not searched for)
- Search Mechanism
- Boolean Searches
- Wildcards
- Other Reserved Characters
- Field specific search: WITHIN Operator
- Date searching
- Non-English characters
- Choosing the display options
- Gisting
- Relevance
- Searching and viewing the results
- Highlighting
Introduction to the Simple Search
The Simple Search interface is available to all users who would like to find information quickly within CORDIS. This includes the CORDIS databases, Web pages and documents. All the user needs to do is to enter one or more specific search terms or words and to click on the "search" button. A results list displays all records matching the search terms – individualrecords can be selected by clicking on an entry from this results list. If more than one search term is entered, an “ AND” operator is assumed between the different search terms. As more terms are entered, the scope of search is narrowed.
topGetting to know the search screen
The design of the simple search screen is focused on meeting the needs of both novice and experienced users. On one page, the user is able to select the type of content they are looking for, enter the desired search terms, control the ordering with which results are displayed, and select the number of results displayed per page.
The following content types have been established: News, Events, Funding, Projects, Partners, Documents, Programmes, Acronyms, Contacts and Exploitable Results.
Finding your way around the search screen:
| 1. Language switching for the search screen |
|
| 2. Database or field of interest selection | |
| 3. Search terms entered by the user | |
| 4. The display options for the results list (standard or detailed, and the number of results per viewable page) | |
| 5. The sort order for the results list | |
| 6. The search and clear buttons for executing a search
or clearing the last actions |
Quick Search (search and go)
The simplest way to conduct a search is to enter one or more words (sometimes known as keywords or search terms) and to click on the search button. (NB: The “enter” key will initiate the search in some browsers, but not all.) The searches are not case sensitive and will ignore the use of upper or lowercase in the searchterms. The number of records returned is limited to 200. If the search query retrieves more than that, the database returns the 200 records with the highestrelevance.
Examples of quick searches are:
| Nanotechnology | (retrieves records that contain the term nanotechnology) |
| Structural Biology | (retrieves records that contain both terms) |
| "Structural Biology" | (retrieves records that contain the phrase "Structural Biology") |
A results list, an error message or a “no records” page will be displayed depending on whether the search is well formulated and has been successful. If a search does not retrieve any records it may well imply that there are no records in the database matching the search term(s). Alternatively there may be spelling errors in the search terms or ambiguous syntax that is interpreted unpredictably. Where possible, the syntax checking will warn users when there are syntax errors.
When there are records displayed in the results list, a simple click on the title will enable the user to view and read the full record. More information on the different options for displaying the results list and the record display are provided below.
topComplex search using the simple search screen
So far, we have introduced the basic searching concepts that are fundamental to the Simple Search; however there are a number of other features including: a full range of Boolean operators for combining search terms; the introduction of special characters to complete pattern matching, and the concepts of Gisting and Relevance. This section will explain these features in further detail.
topFormulating a search
(Note: in this help file, quotation marks ‘ ’ are used to indicate user input unless specified otherwise.)
A simple search term could be expanded in more than one way to make the search as wide as possible. For example:
- A search term like ‘GSM’ would probably return more records
if its expansion was also to be used e.g. ‘GSM or Global System for
Mobile Communications’.
- A search term like ‘Communications’ could be truncated to ‘communicat*’ so
that all possible combinations are retrieved e.g. communicate, communicating,
communication, communications, communicator, communicators, etc.
When entered in quotation marks "" the search term ‘ "Global System for Mobile Communications" ’ returns the records that contain the exact phrase (including stop words as explained in the Stop words section below). Thus, a smaller number of records is retrieved than when quotation marks are not used.
To summarize:
- Search Terms - These are one or more words
entered by the user as follows: 'information technology'
The system will search for the two words in any order and any distance between the two (i.e. other words or sentences may separate the two words).
- Search Terms (exact match) - To specify a particular
phrase or sequence or words, the words are entered with quotation
marks : "information technology"
The system will search for the given sequence, additional spaces or "white space" characters such as tab will be ignored.
Introduction to Advanced Search
The Advanced Search interface provides the user with more search options for more accurate results.
The principles for searching are the same as the simple search, but the Advanced Search allows you to narrow your search by using specific criteria such as 'Category', 'Programme acronym' or 'Subject index'.
Please note that you can only search one type of information at a time.
The 'Save a Search' option (free registration is required) permits you to receive E-mail Notifications when updates match your search criteria.
topMap-based Search
Overview
The map-based search provides a simple graphic interface that allows a search to be made, for Partners or R&D Projects or R&D Results, on the basis of a geographic area defined by the user. It consists of a 'clickable' map, and a 'shopping basket'-type menu showing the list of selected countries and/or regions.
Technical note
The map-based search facility uses 'Macromedia Flash' technology, and therefore requires the 'Active X' controls and plug-in functions to be enabled in the user's computer [e.g. Microsoft Internet Explorer: in the ' Tools' menu select ' Internet Options...', select the ' Security' tab and click the ' Custom Level...' button, in the ' Security settings' menu, all Active X functions should be set to ' Enable' (or ' Prompt' – in which case you will need to manually accept each new Active-X function)]. Certain 'Pop-up' blocking software tools may also inhibit the operation of the map-based search.
Presentation
The initial page of the map-based search function is set out in three 'zones':
- A 'clickable map' of Europe. Only the countries shown on this map are included in the map-based search. For other countries, you may search using other CORDIS search facilities, for example the 'Advanced Search' function, available in the general ' Search' page.
- A zone with the menu for choosing 'Partners' or 'Projects' or 'Results', the list of selected geographic areas (countries and/or regions), and a text-box in which additional search terms can be added if required.
- The 'Submit' button, which will initiate the search you have defined.
Depending on the screen-size setting on your computer, you may be able to view all of these on your screen at the same time, or you may need to 'scroll' down or across to view some of the display. If you cannot see any of the map and other zones, you may need to adjust the settings of your computer (see above technical note).
Selecting the geographic area for searching
As you move the cursor over the surface of the map, the name of the country you point to will appear at the bottom of the map (in the language of the country concerned), and the pointer will change to indicate a clickable function (usually a pointing hand symbol). Click to select a country; a menu will appear at the top of the map offering the choice of selecting the whole country (all regions), or to view a map of the regions in this country. The square symbol in the corner of this menu window will open or close this two-choice menu.
- If you click on ' Add this country to your query', the name of the country will appear in the menu box in zone 2. You have selected all regions in this country.
- If you click on ' Zoom to this country's regions', the map will change to show a map of the regions in this country. Where appropriate, this will include regions outside the geographic zone of Europe – for example the French regions in the Caribbean, or Portuguese Atlantic islands. The name of each region will appear as you place the cursor on it. Clicking on any region will select the region, and its name will appear in the zone-2 menu: you may select multiple regions in the same country. The map menu will offer you the options to still select the whole country, or to ' Go back to the map of Europe'
When you have selected either the whole country, or one or more regions within this country (your choice will be displayed in the menu-box in zone 2), you may return to the map of Europe by clicking ' Go back to the map of Europe'.
You may then choose another country (or regions within this country) to add to your 'menu-box' in zone 2. In this way, by repeating the country/region selection process, you can build up a geographic area corresponding to your needs (e.g. all regions with an Atlantic coastline). Your selected area is listed in the zone-2 menu-box.
Choosing which category of data to search
In zone 2, click on one of the 'radio buttons' to select 'Partners' OR 'Projects' OR 'Results'. Please note that only one can be chosen, you cannot search simultaneously in two or three of these categories.
Review of your geographic area
The 'menu-box' will display the list of countries and/or regions you have selected. If necessary you can remove any selected country or region: click on the item in the menu and then click on ' Remove item'. Multiple-selection is not possible: to remove several items, repeat the individual select/delete action. You can empty the whole list by clicking on ' Remove all'.
Extra search terms
If required, you can provide additional search terms by typing them into the text-box provided. Just adding a word will limit the search to records in the category and area you have defined that also contain this word in any field. Boolean operators can be used.
The 'WITHIN Operator' allows you to define one or more database fields in which the specified search term should appear. For example, if you were looking for a university as a partner, you might specify: 'WITHIN ORGANIZATION_NAME universi*'. It is possible to specify multiple terms, and to edit these within the text box – e.g. copy, paste, delete.
Launch the search
When you have defined the category, geographic area and optional search terms that correspond to your needs, click on the 'Submit' button in zone 3. The animated symbol and text 'Search in progress' will indicate that the system is conducting the search you have defined.
Search results
The results of your search will be displayed as a table of records that correspond to your search. To view a particular record, click on the row of the table: the corresponding data record will be displayed in a new browser window.
The table of search results can be sorted: click on the heading of any column to sort the table in order according to this column. The small triangular symbol in the sorted column header allows you to reverse the sort order – for example A-Z or Z-A, most recent first or oldest first.
The 'Return to map' button will allow you to go back to your defined search in order to modify any of the parameters – for example if you want to narrow or broaden the geographic area defined – and to re-run your search.
topSTOP words (words that are not indexed)
Stop words are words that are not indexed by the system and will therefore be ignored - they are not 'searchable'. These are words such as adverbs, conjunctions, and prepositions, which occur very frequently in text. The idea of excluding stop words is to save space and speed up the indexing and search time. In CORDIS, the following STOP words have been excluded
| a | be | had | it | only | she | was |
| about | because | has | its | of | some | we |
| after | been | have | last | on | such | were |
| all | but | he | more | one | than | when |
| also | by | her | most | or | that | which |
| an | can | his | mr | other | the | who |
| any | co | if | mrs | out | their | will |
| and | corp | in | ms | over | there | with |
| as | for | into | no | so | this | up |
| at | from | is | not | says | to |
The STOP words ' AND', ' OR' and ‘ NOT’ included above are Boolean operators and are not searchable. Their usage is explained below under Boolean Searches
Similarly, the same principle has been applied to STOP words in other languages where there are records in these languages in the database, for example, NEWS, EVENTS, FUNDING and PROGRAMMES.
For example; a search with these two search terms:
‘after because’ retrieves no records because neither of these words are indexed
These three searches have the same meaning because the stop word “for” is removed and AND is assumed for multiple search terms:
‘homepage for Technology’
‘homepage Technology’
‘homepage
AND Technology’
To search for a particular phrase which includes stop words, the search terms must be placed within quotation marks "" for example:
“homepage for Technology”
retrieves only records containing that exact phrase. In this case the ‘for’ is not ignored. For the stop words to be recognized they must be included with non-stop words either side. Adding quotation marks to a series of stop words will not retrieve any records.
topSearch Mechanism
The search mechanism is based on complete words. A word in this case is any string of alphanumeric characters, a to z and 0 to 9 delimited by a space, white space character, punctuation e.g.: -, _ . So as well as normal words, a number like 2003 is also a “word”. When searching, the terms input by the user are compared to the words stored in the index. As explained above, stop words are not loaded to the index and are removed from the user's search terms unless they are part of a phrase within quotation marks "".
The search does not match part words. For example a search for ‘org’ would not retrieve records containing 'organization'or 'organizing'. To achieve this result requires the use of wildcards explained below under Wildcards.
topBoolean Searches
In the simplest case, a search argument can be a single term. However, more than one term is allowed as long as it is separated by at least one space. In this case, the space is interpreted as ' AND' linking the various terms. The words ' AND', ' OR' and ' NOT' have special meaning as they correspond to Boolean operators rather than words to be searched. They are not case sensitive, ‘ and’ and ‘ AND’ have the same meaning.
For example:
‘innovation
AND technology’
‘innovation and technology’
‘innovation technology’
are considered the same and retrieve records containing both ‘innovation’ and ‘technology’
The precedence of the Boolean operators is NOT, AND, OR i.e. the NOT is executed first, then the AND, followed by the OR.
The AND operator
When two or more search terms are linked by the AND operator, the search is for all records containing both terms, e.g.: ‘information AND technology’
The OR operator
When two or more search terms are linked by the OR operator, the search is for either one or the other (or all) terms e.g.: ‘information OR technology’
The NOT operator
In Boolean logic, NOT is a unary operator, acting only on the term that follows it. It is most frequently used with the AND operator. When two or more search terms are linked by AND NOT, the search retrieves records which contain the first term but do not contain the other term(s), e.g.: ‘information AND NOT technology’ retrieves records containing the term ‘information’ that do not contain the term ‘technology’. When NOT is used by itself e.g.: ‘information NOT technology’ an AND is assumed so that the search is the same as: ‘information AND NOT technology’
Parentheses
It is possible to develop a simple search into a complex search where any number of Boolean operators can be used. Parentheses can be used (and we strongly suggest it) to control the sequence and nesting the Boolean operators irrespective of their strength. In this case the precedence is (), NOT, AND, OR.
Examples of search terms using Boolean operator and parenthesis:
‘information
AND technology’
‘information
OR technology’
‘information
NOT technology’
‘(information
AND (technology
OR techniques))
ANDNOT science’
Wildcards
A search term may contain the special characters '_', '?', '$', '*' and '%' - known as wildcards - where either of the first two matches one character, and either of the last three matches multiple (zero to many) characters.
Examples:
| 'Organi?ation' | (retrieves the spelling variants organisation and organization) |
| 'For_' | (retrieves records with the words fora, ford, fore, fork, form, fort) |
| 'Forecast$' | (retrieves records containing the words forecast, forecasted, forecaster etc…) |
| 'Forecast*' | (same as above) |
| 'Forecast%' | (same as above) |
A more complex search example using Boolean operators, parenthesis and wildcards would be as follows:
‘((Diss* AND exploit% AND RT_ AND knowledg?) AND “amended proposal”) ANDNOT (colo_r OR organi?ation)’
topOther Reserved Characters
Several other characters have a special meaning for the databases, these are interpreted by the search interface so that users do not need to take care of them.
topField specific search: ‘WITHIN’ Operator
A new operator has been introduced to enable the simple search user to search in specific fields of a database. This is the WITHIN operator and is not case sensitive.
- The fields that have been made available for this feature are listed below.
| NEWS database | |
| Field Name | Description |
| CATEGORY | Category of News Article |
| CONTACT | Contact Person |
| DATA_SOURCE_PROVIDER | Data source provider |
| DOCUMENT_REFERENCE | Document Reference |
| GENERAL_INFORMATION | Main text of the News article |
| PROGRAMME_ACRONYM | Programme Acronym |
| RECORD_CONTROL_NUMBER | Record Control Number |
| RELATED_NEWS | Related News item |
| CLASSIFICATION | Subject Index codes |
| TITLE | Title / Headline |
| QUALITY_VALIDATION_DATE | Last approval date |
| CLOSING_DATE | Closing date |
| EVENT_DATE | Event date |
| PARTNERS database | |
| Field Name | Description |
| TITLE | Collaboration Title |
| COLLABORATION_DETAILS | Collaboration details |
| EXPIRY_DATE | Collaboration Expiry Date |
| CONTACT_PERSON | Contact person details |
| KEYWORD | Organization Keywords |
| ORGANIZATION_CITY | Organization city |
| ORGANIZATION_COUNTRY | Organization country |
| ORGANIZATION_NAME | Organization name |
| PARTNERS_ACQUIRED | Partners already acquired |
| QUALITY_VALIDATION_DATE | Quality validation date |
| RECORD_CONTROL_NUMBER | Record control number |
| ORGANIZATION_DETAILS | Organization details |
| CLASSIFICATION | Subject Index codes |
| CATEGORY | Category |
| PROGRAMME_ACRONYM | Programme Acronym |
By entering a search term and selecting where to search, the user will retrieve more accurate results.
| Examples: (News database) | |
|
|
|
A more complex search example using Boolean operators, parenthesis, wildcards and the WITHIN operator would be as follows:
‘(((Diss* AND exploit% AND RT_ AND knowledg?) AND “amended proposal”) ANDNOT (colo_r OR organi?ation) WITHIN title)’
topDate searching
The date format used throughout CORDIS and recognized by the searches is YYYY-MM-DD. Date searching in the simple search has several restrictions. Date searching cannot be combined with comparison operators such as ‘=’, ‘>’, ‘<’, ‘>=’ or ranges of dates. These features will be available for searching in the advanced and professional searches.
Examples of date searches:
| ‘2003-11-20’ | (retrieves records that have the date 2003-11-20 in text or date fields) |
| ‘2003$’ | (retrieves records that have a date of 2003 in text or date fields) |
| ‘2003-03$’ | (retrieves records that have a date of 2003-03-dd in text or date fields) |
The date search can be restricted to a particular date field using the WITHIN operator explained above Field Specific Search. For example in the Partners database, a search with:
‘2003-11-20 within EXP_DATE’ retrieves records that expire on 2003-11-20, and not records that happen to have this date elsewhere in the text.
topNon-English characters
All characters are treated literally. To search for a word containing, for example, a French character with an accent, requires entering the word with that character. For languages such has German there may be alternate official spellings of a word, these will not be recognised at present, but we plan for such terms to be recognized in future.
topChoosing the display options
- Results. The results list may be standard or detailed: i.e. in the standard option the most important fields are shown, whereas the detailed option contains a little more information on the record (e.g. a short extract of the text field or more descriptive information on the record
- Sort Options. These are primarily based on date and relevance. In the results lists, more sorting options are offered, but not all of the fields are available for sorting.
Gisting
Gisting is a feature that automatically selects a paragraph that best represents the record returned. The feature is configured to search the main text fields of the record and construct an extract from them. The resulting summary is presented in the 'Detailed' results list to give the user an idea of the content before retrieving the entire record. The detailed results list is obtained by selecting 'Detailed Format' on the search screen before initiating the search.
topRelevance
CORDIS has introduced the feature of relevance ranking in its searches. To
calculate a relevance score for a returned document in a search term query,
CORDIS uses an inverse frequency algorithm based on Salton’s formula.
Inverse frequency scoring assumes that frequently occurring terms in a document
set are “noise” terms, so these terms are scored lower.
For a document to score high relevance, the search term must occur frequently
in the document but infrequently in the document set (i.e. the list of returned
documents corresponding to the search) as a whole.
In the results list, the score displayed is not a percentage. For example: supposing that in a document set:
- There were 200 documents dealing with chemistry in which the term ‘ chemical’ occurs at least once in every document. The term ‘chemical’ thus occurs frequently in the document set.
- A document contains five occurrences of ‘ chemical’ and five occurrences of the term ‘ hydrogen’. No other document contains the term ‘hydrogen’. The term ‘ hydrogen’ thus occurs infrequently in the document set.
Since ‘ chemical’ occurs so frequently in the document set, its score for the document is lower with respect to the score for ‘hydrogen’, which is infrequent in the document set as a whole. The score for ‘ hydrogen’ is therefore higher than that of ‘ chemical’. This is so even though both terms occur five times in the document.
Inverse frequency scoring also means that if CORDIS were to add documents that contain ‘ hydrogen’ this would lower the score for that term in the document, and adding more documents that do not contain ‘ hydrogen’ raises the score. Thus, scores for the same document can vary from day to day as documents are added.
Note: Even if the relatively infrequent term ‘ hydrogen’ occurred four times in the document, and ‘ chemical’ occurred five times in the document, the score for ‘ hydrogen’ might still be higher, because ‘ chemical’ occurs so frequently in the document set (at least 5 000 times). Searches using quotation marks (e.g. “information technology”) will return a higher relevance score for terms that appear in the same order. A lower score will be given if they appear in any order.
topSearching and viewing the results
To execute the search, the "Search" button needs to be clicked. (NB: The “enter” key will initiate the search in some browsers, but not all.) There are three possible outcomes:
- A successful search with a maximum of 200 records; a results list will be displayed. When the search terms match more than 200 records, the 200 with highest relevance are returned. This limit is enforced to prevent the server resources being taken up with inefficient searches. The number of results displayed per page (10, 20, 30, 50, 100, - default value is 10) may be set on the search screen or on the result list page. The result list may be sorted by relevance, date, title etc.
- An unsuccessful search ("Search Failed") where no records were found matching the search string or criteria entered. This does not necessarily mean that the search string is incorrect. There could be no records found, there could be a spelling error entered in the free-text fields, incorrect use of Boolean operators, etc.
- A Syntax Error in the search terms; a warning message is displayed.
The titles in the results list are hyperlinked to the actual record where it can be viewed. Within each record, there are various links to related records in other databases, which are embedded within the record and link to general information such as Programme Acronyms, etc.
The full record may be printed from the results list.
topHighlighting
Highlighting is a useful feature that enables search terms to be more prominent in the results lists and record display. A search term like ‘CORDIS’ will be highlighted as CORDIS whereas a search term like ‘org*’ will have all matching terms highlighted for example: organ, organize, organization, etc.
Last updated on: 2008-03-26
