Attribute coding setup Example

System Setup

When you first get your Influence Professional database you will want to setup an Attribute structure to allow you to 'tag' candidates with particular skills and attributes that you can later use to search for these candidates.

The purpose of this attribute (coding) system is to allow you to quickly record some of the KEY skill areas or abilities that candidates have, and it is not necessary to create every search term you will ever use, since you can easily search for more niche items using the CV Text Search facilities.

Initially you should try to think about major Categories of skill, and then find particular terms within those categories. It is sometime useful to use MS Excel to build a replica of the skill structure that you wish to use so that you can picture what you are creating. You are attempting to create a 'Tree' of attributes that allows you to Quickly & Easily classify candidates - hence why you do not want to make this too complex at this stage.

Example Attribute (Coding) Structure

If you were using the database for recruiting into the IT industry you may have categories and items as below:
Databases
       Oracle
       SQL  Database                      << In this example SQL has been sub-divided into MY sql and MS sql.
                          MY SQL
                          MS SQL/SQL Server
        Informix
        DB2
Programming Languages
         C#
         .Net
         HTML
         Java
Web Development
         SEO
         E-Commerce
         Content Management
         Actionscript
         PPC (Pay-per-Click)
         Database Optimisation
Personal Qualities
        Good Personality
        Works well with Teams
        Self-motivated  

Entering the Attribute (Coding) Structure

Once you have decided upon a STRUCTURE for your attribute coding system (or KEYCODES) you need to enter this structure into the Influence Professional database. This entry can only be done by Administrator users within the database or those with Administrator privileges.

Choose Maintenance > Setup > Attribute Maintenance.
Once the screen appears you may click the [         + ] button to add a new item at any level. Type the description of the item, then press <Tab> to move to the shortcode and press <Tab> again to complete the entry.

e.g. To create the initial categories above click the [    +] at Level 1 and enter Databases, Programming Languages and Web Development

To create the items at 'Level 2' simply select one of the categories you have just created and right-click on the item and choose 'Add item below'. 
This will allow you to create the actual attributes (e.g. Oracle, SQL, Informix, DB2) - repeat the process for each Cateory (Level1) item you have created. For items with a deeper structure (eg SQL) you can repeat the process of right clicking to create additional levels.

Optional: 
Once you have created your Attribute coding structure, you can specify certain KEYWORDS in a CV which will cause the system to auto suggest that attribute when a CV is imported into the system.
To do this simply click the + in the 'Pat' column next to the attribute you wish to amend. Then use the Word Patterns screen to enter the combinations of Words/Phrases on a candidate CV that should be used to suggest they posses this Attribute.
e.g. To suggest the Attribute Actionscript you may tell the system to look for
Actionscript     <OR>
Action Script

Note: If you do not setup any word patterns then the system will never suggest this attribute for a candidate and the attribute must always be manually chosen if you wish a candidate to have it recorded.

Setting Automatic Word Patterns

It is possible to setup the attributes system so that CVs are ‘scanned’ looking for word patterns. If the word pattern is found the system will allocate the attribute as Non Validated (i.e. shown as ƒ rather than P or Ï )

Note: When using <NEAR> the system will only consider the item immediately below. i.e. To search for French near Fluent or Speaking, you need to setup French <near> Fluent or French <near> Speaking

You can specify the number of ‘Hits’ before an attribute is allocated by using {{xx}}

In this example the system would need to see the phrase ‘TB’ five times before it would allocate the attribute.

Or Trial Balance just once.

Examples of Word Patterns

Shown below are examples of how to setup the word patterns to achieve various different results when highlighting the text of CVs.

Wildcard (*)

The * can be used as a wildcard when creating patterns. For example WELD*, will pattern match with WELD, WELDER, WELDING, WELDED.

Pattern	Logic	Dist
WELD*	OR
TIG	OR
MIG	OR
ARC	OR

Groups 

When searching for patterns it is useful to GROUP terms together using [ and ].

This example would look for the word Oracle and any one of Money, Salary or Financ* (i.e. Finance, Financial, Finances, etc).

Pattern	Logic	Dist
ORACLE	AND
[Financ*	OR
Money	OR
Salary ]	OR

NOT

It is possible to specify words that must NOT be found by using ~ to mean NOT

E.g. This will allocate the attribute if the word CIDER is seen at all, or the word “PERRY” as long as FRED is not seen at all.

Pattern	Logic	Dist
PERRY	AND
~FRED	OR
CIDER

Near 

When setting up ‘Near’ the system only looks at the row IMMEDIATELY below the word.

This example would look for Dinosaur near fossil or Dinosaur near Bones

Pattern	Logic	Dist
DINOSAUR	NEAR	10
FOSSIL	OR
DINOSAUR	NEAR	15
BONES

**IMPORTANT** If at all possible it is best to avoid the use of the <NEAR> option unless absolutely necessary. It is much more complex to calculate and over-use can affect system performance.

TIP:  Try to avoid the use of ‘common’ words such as ‘IT’  - you may mean ‘Information Technology’, but the system will look for the word ‘it’ – which will be wrong