WO2013045949A1 - Requirements characterisation - Google Patents

Abstract

A method of determining a requirements characterisation profile for an entity is disclosed. The method comprises the steps of receiving classification parameters defining a requirement for an entity and selecting, in dependence on the classification parameters, a set of entities from a database of previously assessed entities. The method further comprises retrieving from the database characterisation parameters of the selected set of entities,and constructing, in dependence on the characterisation parameters, a requirements characterisation profile for the entity. An apparatus is provided that comprised means for performing such a method.

Description

Requirements characterisation

This invention relates to apparatus for and method of providing a requirements characterisation profile for an entity. In particular, the invention allows for the translation of a generic requirements request into a specific requirements request. Variants of the invention may allow for translation between different models of requirements between different organisations. The invention may also allow for the review and revision of the resulting requirements request. Embodiments of the invention may also provide recommendations of suitable assessments for determining whether the determined requirements are met. The invention has relevance to a wide number of fields in which a user is required to draw up a requirements request based on an apparently limited amount of relevant initial information, for example when a requirements request is required to be drawn up by a non-specialist. A specific example is presented in the field of human resources, describing how such a system may assist in drawing up a job description for a job vacancy. A common problem in many fields is that a requirements request - an order for a part, a request for tenders, even a job description - is often not fit for purpose. The request may be unhelpfully generic, resulting in many clearly unsuitable matches needing to be filtered out. In some cases, the request may be little more than a "wish list", specifying features which may have little bearing on what is actually required to fulfil the requirement. For example, features may be described as "essential" which in reality are not, resulting in potentially suitable matches being needlessly rejected. In other situations the request may fail to list (or give insufficient weighting to) important features, resulting in matches that later prove to be unsuitable.

These problems clearly result in inefficiencies, but in certain circumstances could also result in more serious consequences. For example, a request for a replacement machine part based on incomplete or inappropriate data which may result in an unsuitable part being used, and the part potentially failing catastrophically. Such a situation could arise, for example, when parts exactly matching the original are no longer manufactured and it may not be known precisely which features of the original part are essential and which could be approximated or ignored, without compromising performance or safety (but potentially resulting in a cost saving).

As a further example, in the field of human resources, a common complaint is that the hiring process is often imprecise and arbitrary. Typically, a manager (such as a line manager or human resources manager) decides themselves what features of a candidate are considered to be relevant for a particular job vacancy, draws up a corresponding job description comprising a list of required skills, experience and qualifications, and assesses candidates accordingly for the presence of those required skills, experience and qualifications. In some cases, a job specification would also list desirable competencies of a candidate. Invariably, however, the existence of a very wide range of possible requirements means that these lists of requirements are often poor tools for determining candidate suitability. This is perhaps unsurprising, given that those drawing up the hiring specification are usually unlikely to have exhaustive knowledge of all of the required skills, experience, qualifications and competencies deemed necessary. A poor job description is therefore often the weak link in the hiring process - and this may be true even if the subsequent assessment is entirely valid; the wrong competencies may be being assessed. The result may be the placing of candidates in unsuitable jobs, as well as jobs being filled by unsuitable (or at least non-optimal) candidates. Generally, these situations may arise because of lack of knowledge and/or experience of the individual (perhaps a non-specialist in the area) tasked with drawing up the requirements request. Alternatively, there may simply be a lack of relevant information on which to base the requirements request.

The present invention aims to address at least some of these problems.

According to one aspect of the invention, there is provided a method of determining a requirements characterisation profile for an entity, comprising the steps of: receiving classification parameters defining a requirement for an entity; selecting, in dependence on the classification parameters, a set of entities from a database of previously assessed entities; retrieving from the database characterisation parameters of the selected set of entities; and constructing, in dependence on the characterisation parameters, a requirements characterisation profile for the entity.

Further features of the invention - which may be provided dependency on or independently of the above or any other aspect(s) of the invention, in any appropriate combination - preferably include:

• At least one database, comprising entity data. Preferably, the entity data comprises assessment data (such as characteristics and/or performance data).

• At least one server, for housing the at least one database. A plurality of servers may also be used, for example in a distributed or redundant arrangement.

• A method of receiving an initial requirements request.

• A method of receiving a requirements model.

• A method of receiving a set of classification parameters. Preferably, the classification parameters allow an entity to be identified generally. More preferably, the classification parameters are in the form of a nested list of classes and sub-classes.

• A method of identifying in dependence on the classification parameters a set of potential candidate entities.

• A method of determining a characterisation profile comprising a set of characterisation parameters of the set of potential candidate entities. Preferably, only a subset of characterisation parameters is used. More preferably, only the most significant characterisation parameters are used. The most significant characterisation parameters may comprise those deemed to be most relevant and least relevant. Preferably, the characterisation parameters may be competencies.

• A method of assigning the characterisation parameters to an importance class. Preferably, the importance class may be one of: essential, desirable, relevant, non-relevant. Preferably, the method further comprises determining a characterisation profile or model determined from, preferably comprising a subset of, the characterisation parameters.

• A method of determining the fraction of members of the set of potential candidate entities with characterisation parameters in each of the importance classes. Preferably, the method further comprises determining the probability of a randomly selected member of the set of potential candidate entities having a characterisation parameter in an importance class. More preferably, the method further comprises determining the probability for each of a set of characterisation parameters for each of a set of importance classes.

• A method of receiving a set of context parameters or variables. The classification parameters or variables may be determined from the answers to a series of questions concerning aspects of potential or candidate entities. Preferably, the method further comprises determining modifiers with which to adjust the probabilities in dependence on the context parameters. Preferably, the probability adjustment comprises a Bayesian probability calculation. • A method of determining a relevant set of context parameters. Preferably, the context parameters are determined in dependence on the classification of the entity. The context parameters may be determined from previously determined context parameters. Preferably, the number of context parameters used is in dependence on the classification of the entity. Preferably, the set of context parameters used is those which are determined to have potentially greatest effect on the resultant probabilities. The method of determining whether a context variable is deemed "relevant" may make use of Kendall's Tau-C statistic.

• A method of generating a set of requirements or a requirements request in dependence on the determined probabilities and adjusted probabilities of characterisation parameters and importance classes.

• A method of receiving an adjustment input in respect of the set of requirements or the requirements request. Preferably, the adjustment input relates to assignment of a characterisation parameter to an importance class. More preferably, the method further comprises adjusting the assignment of a characterisation parameter to an importance class. · A method of forwarding the set of requirements or requirements request to a third party and for receiving adjustment inputs from the third party. The adjustment inputs from the third party may be distinguished by a suitable label, including representation of the third party adjustment input with a symbol, icon, identification number, and display of hover-over narrative.

• A method of generating a revised set of requirements or requirements request in dependence on a received adjustment input.

• A method of determining a suitable assessment method or technique in dependence on the set of requirements or requirements request for assessing potential or candidate entities; preferably, a method of selecting a plurality of assessment methods or techniques in dependence on the set of requirements or requirements request for assessing potential or candidate entities. Preferably:

- a plurality of assessment methods or techniques is selected in dependence on a plurality of characterisation parameters such as competencies

- a respective characterisation parameter weighting, or an importance class, is associated with each characterisation parameter, and, preferably, the plurality of assessment methods or techniques is further selected according to the characterisation parameter weighting (or importance class), with, more preferably, only assessment methods or techniques related to higher weighted (or higher classified) characterisation parameters being selected

- the plurality of selected assessment methods or techniques may be limited to a maximum number of assessment methods or technique

- the selected assessment methods or techniques may be optimised to relate to a maximum number of characterisation parameters

- the selected assessment methods or techniques may be optimised to relate to a maximum number of characterisation parameters in a particular importance class (for example 'high' importance)

- the selected assessment methods or techniques may be optimised to relate to a maximum number of combined and weighted characterisation parameters (for example using characterisation parameter weightings)

- a respective assessment method weighting is associated with each assessment method and characterisation parameter

· A method of evaluating a plurality of assessment results to generate a plurality of characterisation parameter results, preferably individual assessment results relating to a plurality of characterisation parameters. A plurality of assessment results may be weighted and combined to generate a characterisation parameter result.

• A method of evaluating a plurality of characterisation parameter results in dependence on a weighting input, said weighting input preferably being said characterisation parameter weighting (or importance class). Preferably, an assessment score (preferably a single assessment score) is calculated in dependence on the plurality of assessment results and weighting inputs.

• Integration of the method in a platform that enables further functionality, for example administration of candidate entities; submission of invitations to candidates to undertake an assessment; administration of assessment results; collecting, scoring, evaluating and/or reporting of assessment results; presentation of assessment results; and/or other functions relating to a characterisation profile and potential candidate entities.

• A method of determining a job specification, comprising: receiving a job classification; selecting from a database in dependence on the classification a set of matching job profiles; extracting for each matching job profile a set of assessed competencies; determining the relative relevance of each of the competencies to the job classification; and generating a job specification based on the determined relevant competencies. This may allow a user who has little understanding of job profiling to produce a set of required competencies that is based on competencies identified as important in previous profiles for similar jobs. It may also allow users to plan for future job vacancies and succession plans.

· Preferably, the job classification comprises a standard classification, such as the Standard

Occupational Classification (SOC) code as used by 0*NET. The classification may be based on a hybrid of standard classifications, for example using the granularity provided by a first classification, in combination with the set of aggregations of levels provided by a second classification.

· Preferably, the competencies are defined according to a competency model of a sufficient number of parameters to allow for compatibility by mapping to an externally supplied competency model. An example of such a model is the Universal competency Framework (UCF) developed by SHL Group.

• The relevant competencies for a particular job may be determined from data compiled from assessments of workers employed in the same or similar jobs.

• Preferably, the relevance of competencies of a job specification may be adjusted according to a set of contextual variables determined by answers provided by the user to questions relating to particulars of the job. The particulars may comprise a requirement of the job, the job environment or of the job candidate.

· Preferably, the relevance of competencies of a job specification may be adjusted by a moderator to which the job specification is referred.

• At least one server, for providing access for a client or user either directly or via a computer, for example via a web interface, to the results of processing.

• One or more server functions may be combined in a single server.

· Suitable computer servers may run common operating systems such as the Windows systems provided by Microsoft Corporation, OS X provided by Apple, various Linux or Unix systems or any other suitable operating system.

• Suitable databases include ones based on SQL, for example as provided by Microsoft Corporation or those from Oracle or others.

· Remote access to the analytics system may be provided via one or more web servers configured to provide a website or other remotely-accessible interface. Web interfaces and other code may be written in any suitable language including PHP and JavaScript. A Microsoft .Net based stack may be used.

A method, preferably as foresaid, further comprising determining a raw profile for the entity in direct dependence on the selection of the set of entities in dependence on the classification parameters.

Preferably, the requirements characterisation profile is available alongside the raw profile.

Preferably, wherein the constructed requirements characterisation profile can be overridden.

Preferably, wherein the requirements characterisation profile is available in adjustable slider format, and preferably raw profile is available in fixed slider, more preferably both in conjunction with each other.

A method of facilitating the recruitment of a job candidate, comprising: receiving job classification information; optionally, receiving job context information; determining, in dependence on the job classification an initial competency profile; and, optionally, refining the competency profile in dependence on the job context information. Preferably, the method further comprises receiving and incorporating feedback on the competency profile. Preferably, the method further comprises determining a suitable assessment for assessing potential candidate for competencies in accordance with the competency profile.

Generally, a method and apparatus is provided for the translation of a generic requirements request into a specific requirements request based on the probabilistic analysis of historic assessment data of a population of entities matching classification parameters determined from the generic request. The term "competency" as used herein may refer generally to any measureable characteristic.

The invention also provides a computer program and a computer program product for carrying out any of the methods described herein, and/or for embodying any of the apparatus features described herein, and a computer readable medium having stored thereon a program for carrying out any of the methods described herein and/or for embodying any of the apparatus features described herein.

The invention also provides a signal embodying a computer program for carrying out any of the methods described herein, and/or for embodying any of the apparatus features described herein, a method of transmitting such a signal, and a computer product having an operating system which supports a computer program for carrying out the methods described herein and/or for embodying any of the apparatus features described herein. The invention extends to methods and/or apparatus substantially as herein described with reference to the accompanying drawings.

Any feature in one aspect of the invention may be applied to other aspects of the invention, in any appropriate combination. In particular, method aspects may be applied apparatus aspects, and vice versa.

Furthermore, features implemented in hardware may generally be implemented in software, and vice versa. Any reference to software and hardware features herein should be construed accordingly. The invention will now be described, purely by way of example, with reference to the accompanying drawings, in which:

Figure 1 shows an overview of a system for providing a requirements characterisation profile for an entity; Figure 2 shows, in overview, the system in use;

Figure 3 shows in overview the system as applied in a human resources "job-matching" application; Figure 4 shows the process of determining an initial entity characterisation profile;

Figure 5 shows a flowchart of the system processes;

Figure 6 shows the login page presented by the system;

Figure 7 shows examples of initial user interfaces presented by the system;

Figure 8 shows an example of an entity classification model;

Figure 9 shows examples of initial specification description screens;

Figure 10 shows the "create new job screen";

Figure 1 1 shows examples of context variable input screens;

Figure 12 shows another overview of the process of competency profile estimation and modification;

Figures 13 and 14 show an example of the calculation of modified entity characteristics profile;

Figure 15 shows in tabular form the evolution of characterisation prior probabilities with successive iterations;

Figure 16 shows in graphical form the evolution of characterisation prior probabilities with successive iterations;

Figure 17 shows an example of a single competency histogram;

Figure 18 shows an example of a full final entity characteristic (competency) profile;

Figure 19 shows examples of an entity characteristics profile summary page;

Figure 20 to 22 show examples of an interactive profile modification facility;

Figure 20 shows a Rate competency importance page;

Figure 21 shows an example of a hover-over / pop-up panel;

Figure 22 shows the Rate competency importance page with adjustment;

Figures 23 to 26 show an external profile feedback facility;

Figure 23 shows a feedback progress screen, also termed a "Rater status" screen;

Figure 24 shows the rater landing page;

Figure 25 shows the rater overview screen;

Figure 26 shows the final profile review summary page;

Figure 27 shows the completion status page;

Figure 28 shows examples of the job library page;

Figure 29 shows an illustration of how the relevance of a context variable is determined; and

Figure 30 shows an example of the selection of assessments and weighting of scores.

Overview

Figure 1 shows an overview of an expert system for providing a requirements characterisation profile for an entity.

The system 1 comprises a server 100 adapted to send and receive data to and from a number of sources over a computer network, including a client terminal 102 from which it receives an initial entity requirements request and to which it transmits a calculated requirements characterisation profile.

Server 100 provides access to a proprietary database 106 of entity assessment data 108. This data comprises the results of entity assessments assembled over many years. Database 106 is also used to store calculated requirements characterisation profiles, preferably at various stages of computation and moderation.

The entities data 108 in the database 106 have previously been analysed in terms of characteristics and are classified according to classification parameters that allow sets of similar entities to be identified. Typically, the entities in the entity database are classified according to a master set of classification parameters sufficient in number to allow mapping to a preferred set of classification parameters, such as that obtained from external server 1 10. Where the externally supplied classification data is of superior quality for some fields it may be used to enhance the classification of the entity database 106.

Server 100 is optionally in communication with moderator terminals 1 14. Moderators 1 16 may be requested to pass comments, adjust or approve the determined requirements characterisation profile.

Server 100 is also configured to determine suitable assessments to be applied to test candidate entities 1 18 according to the calculated requirements characterisation profile.

In an exemplary embodiment, such a system is used in the field of human resources and facilitates the provision of job descriptions. Database 106 comprises a proprietary data set of competency assessments 108 derived from the results of approximately 7,000 unique jobs (the number of analyses increasing as more data is collected), that have been analysed, placed in the database, and mapped to the occupation classification system. Job analysis may be a time-intensive procedure, often requiring several days for a skilled analyst to accurately determine the competencies required for carrying out a particular job. The job analyses are classified according to a set of master classification parameters (typically 100 or more) which are then mapped to a standard occupation classification.

For example, 0*Net is a job analysis database, which makes use of an occupation classification system called "0*Net SOC" to organise its job data. This is a more refined version of the US Government Department of Labor SOC framework, in the sense that at the most detailed level it provides more granularity. However, in the "0*Net SOC" does not make use of all the levels of aggregation provided by the standard SOC. In this embodiment, a modified classification is used, which uses the granularity provided by the 0*Net SOC and the full set of aggregations of levels provided by the standard SOC.

For example, a typical number of uniquely identifiable occupations at each level may be:

• Level 4: 1 171

· Level 3: 449

• Level 2: 96

• Level 1 : 23

Database 106 is therefore based on real-world data and allows for identification of the actual relevance of the various competencies of workers currently (or previously) in identical or similar jobs.

Figure 2 shows, in overview, the system in use.

• Step S1 - User profile requested

User 104 selects the appropriate entity classification on terminal 102 from a drop down or cascade menu 200.

• Step S2 - Initial profile determined

Server 100 receives the selected classification from terminal 102 and looks-up entities 108 matching the classification from the entity database 106. In order to maintain statistical significance, a threshold minimum number of entities (dependent on the level of entity classification selected) is required to fall within the final classification, else the above classification is used. A representative set of defining characteristics is selected as the basis of the entity profile or fingerprint, and the relative proportion of entities matching the classification having the characteristic to specific extents is determined, thereby indicating whether each characteristic is - in order of relevance - "essential", "desirable", "relevant" or "non-relevant". This results in an initial or starting entity profile, essentially a distribution of the relative importance of the various characteristics. The profile is stored in the entity database.

• Step S3 - Profile modified (Optional)

If the user wishes, the entity profile may be subsequently modified. One way of achieving this is by the use of context variables 202, determined by interaction with the user to establish further details of relevance. These context variables are used to modify the competency profile by means of Bayesian methods. Working with probabilities in this way allows for this new data to be taken into account, whereas traditional "frequentist" statistical methods applied to the entity database would be constrained to the frequency distribution of the existing data.

• Step S4 - Profile finalised

The final entity profile relates the relevance of entity characteristics to the initial entity classification, for example shown as a histogram 204 detailing for each characteristic whether it is deemed "essential", "relevant", "partially relevant" or "not relevant". The profile is stored in the entity database.

The process by which an entity profile is calculated from the initial classification and the subsequently determined context variables is discussed in more detail below.

The resulting estimated characteristics profile can be used for example to draw up a specification request and subsequently used for matching to possible candidate entities. Alternatively, it can be used to inform further decision making regarding desired entity characteristics. The profile can also be used for benchmarking purposes.

Figure 3 shows in overview the system as applied in a human resources "job-matching" application. The steps are analogous to those previously described.

• Step S1 - User profile requested

The user broadly identifies job from the "standard occupational classification" or SOC code by selecting from the options presented at each level of the SOC hierarchy 300 using drop-down menus.

· Step S2 - Initial profile determined

Job profiles within the chosen occupation classification are looked-up in the database (these typically number 30-40, but could range from a minimum of 16 to several thousand) and an occupation profile (based on representative competencies) for the selected job classification is calculated. This forms an initial occupation profile, detailing which competencies are deemed to be important for the job in question.

· Step S3 - Profile modified (Optional)

Here, the modification process comprises defining the job in more detail 302. The user is asked approximately 20 questions (answers to which are selectable from drop-down menus) relating to job and the job environment, for example relating to the size of the organisation, specific educational requirements etc. The context questions asked are dependent on the occupation that has been selected. Analyses have been conducted to determine which set of context questions are most likely to influence importance of competencies, for each occupation. This information is used to adjust initial competency profile. For example, it may be known from the data that a particular competency, such as initiative, may be more important in smaller organisations.

• Step S4 - Profile finalised

The resulting profile is a recommendation of relevant competencies for the job.

In order to maintain statistical significance, a threshold minimum number of jobs is required to fall within the final classification, else the above classification is used. This minimum number of jobs required for the selected occupation varies depending on the level of occupation selected. For example, the minimum required number of jobs may be:

Level 4: 40 jobs

Level 3: 40 jobs

Level 2: 20 jobs

Level 1 : 16 jobs

In this embodiment, the user does not have the option to "modify" the profile by first inspecting the initial "prior" profile based on the occupation only, and then providing context variables to adjust the predicted profile. Rather, the user is required to provide the occupation code and has the option to provide one or more pieces of context information. The user then receives a predicted profile based on the input provided. Generally, modification of the profile may refer to any changes the user makes to the profile away from the predicted profile. For example, the system may estimate that competency X is "A key part of the job", but the user can choose to classify competency X as "Not part of the job".

The resulting estimated competency profile can be used to draw up a job description and subsequently used for matching to job candidates, for example by identifying the best matched candidates for a job. Alternatively, it can be used to inform further decision making regarding the job competencies, for internal selection, promotion, succession or outplacement. The profile can also be used for benchmarking employees and organisations. The job profiles created could also be used to generate a shortlist of behavioural descriptions that can be used within a 360 feedback system.

Figure 4 shows the process of determining an initial entity characterisation profile. A set of entities 406 is selected from the entity database 106 based on matching to the cascaded entity classification (higher level classifications: 402 and 404) submitted by the user. For each of a set of characteristics (determined as those which are most influential), the possible characteristic assessment results are divided into four groups - corresponding to very high (essential) 408, high (relevant) 410, low (partially relevant) 412 and very low (not relevant) 414 values - and the fraction of entities determined in each group. This results in a probability distribution for each characteristic, indicating from real-world data the significance of the particular characteristic for entities known to belong to that classification. Also shown is the use of a context variable 416. Once an occupation has been selected, the number of jobs the estimation is based on is the number of jobs in the database mapped to that occupation code. This full set of jobs is drawn on for each context variable used. The only exception is rare instances of missing data i.e., an analysed job in the database might not provide a value for a context variable). A revised conditional probability is calculated for the probability distribution for each characteristic, taking the factor represented by the value of the context variable into account.

Figures 5(a) and (b) show a flowchart of the system processes. Figure 6 shows the login page presented by the system.

Figures 7(a) and (b) show an initial user interface presented by the system, in this example, for a "job- matching" application. Various options available to the user (not all are shown) include: ^■ create a new profile 700

^■ view existing profiles

^■ use an existing profile

■ modify an existing profile

^■ forward a profile for comments

^■ invite rating of a profile 706

^■ view available assessments

^■ view assessments recommended for a profile

■ create a job description

^■ forward a job description to a potential candidate

^■ view candidates

^■ view assessment completion status 708

^■ buy credits 710

■ use unused reports 702

^■ use extended reports 704

Figure 8 shows an example of an entity classification model. This example is based on the Standard Occupational Classification (SOC) model used by the US Department of Labor with additional features from the 0*NET SOC. The classification model is hierarchical, comprising four main levels of increasing specialism:

^■ Major group 800

^■ Minor group 802

■ Broad occupation 804

^■ Detailed occupation 806

Individual job titles are mapped into the most detailed level - "Job titles" (approximately 1000 in total) 808 - although in this embodiment this level is not made available to the user. This allows all jobs to be classified into generic occupations, as used in the 0*Net occupation database. For example, major groups (23 in total) comprise the following:

1 1-0000 Management

13-0000 Business and Financial Operations

15-0000 Computer and Mathematical

17-0000 Architecture and Engineering

19-0000 Life, Physical and Social Science

21-0000 Community and Social Services

23-0000 Legal

25-0000 Education, Training and Library

27-0000 Arts, Design, Entertainment, sports and Media

29-0000 Healthcare Practitioners and Technical

31-0000 Healthcare Support

33-0000 Protective Service

35-0000 Food Preparation and Serving

37-0000 Building and Grounds Cleaning and Maintenance

39-0000 Personal Care and Service 41-0000 Sales

43-0000 Office and Admin Support

45-0000 Farming, Fishing and Forestry

47-0000 Construction and Extraction

49-0000 Installation, Maintenance and Repair

51-0000 Production

53-0000 Transportation and Material Moving

55-0000 Military Specific The use of SOC is provided as an example of an entity classification model; generally, a hierarchical method is useful for quickly finding a job in a library of jobs.

Figures 9(a) and (b) show examples of initial specification description screens, in these examples, for a "job-matching" application. Standard text input fields allow for job title 900 and description 902 to be input; optional job information fields 904 allow for additional text for inclusion in the Job profile report. Occupation drop-down menus 906 allow the user to select the correct SOC code. User inputs are saved when upon completion the user navigates to the next page, otherwise a warning message reminds the user any data they have input will be lost. The can subsequently return to this page to make additional changes.

Figure 10 shows the "create new job screen".

Figures 1 1 (a), (b) and (c) show examples of context variable input screens, in this example, for a "job- matching" application, allowing additional job information can be input. An array 1 100 of approximately twenty questions is presented, the answers to which (selected by the user via the drop down menus 1 102) are set as context variables. Responses to these questions are not compulsory. Users can make no changes and click 'Create New job' 1 104 if they wish to. As the use of context variables may allow for important modification of the initial characteristics profile, this is a charged service. In some embodiments, the number of context variables used and/or displayed to the user is determined by the system. This may be based, for example, on their relevance and impact for the specific job or job category selected. In other words, the system decides which context variables will materially change the competency weightings if they are adjusted (meaning which ones will make a material difference to the job) and only displays those to the user.

For example, in some cases, the user will only see ten context variables for the job they have chosen, as the system has calculated that the other twelve do not make any material impact to the competency weightings whichever values are chosen. In another example, the user will be presented with twenty context variables since they could all have an impact on the competency weightings.

The advantage lies in not wasting the user's time trying to find information that will have no impact on the competency weightings.

The generic idea is that of identifying the most relevant context variables for an occupation in a data- driven manner. The user is free to modify the entries and selection on the initial specification description and context variable input screens until ready to proceed. Once use of this feature is accepted (by selecting "create new job"), a charge unit is deducted (credit allowing) from the user's account. Figure 12 shows another overview of the process of competency profile estimation and modification, in this example, for a "job-matching" application.

• At Step 1 , the user selects the relevant SOC classification and the system evaluates an initial competency profile, by calculating the fraction of jobs assessed at each competency level. This allows for initial statements to be made of the form: "In n% of jobs like yours, competency X is a key part of the job".

• At Step 2, the user sets a first context variable (in this example "Education" to "... requires a college degree"). The system recalculates the competency profile accordingly, taking this context variable into account.

· At Step 3, the user sets a further context variable (in this example, "Size of organisation" to

"1000-2000 employees"). The competency profile is recalculated, taking this further context variable into account. The process is repeated for each context variable set.

• At the end of the process, a final estimated profile is determined. This final estimated profile is displayed a graphical format, as a histogram 1800 and as a table 1900. A user-editable form may also be presented.

In this embodiment, the user only sees the final profile, as it stands after taking into account the selected occupation and any context questions (variables); in alternative embodiments, the effect of each context variable on the profile is shown separately, for example as each context variable is input and its effect determined.

Figures 13 and 14 show an example of the calculation of a modified entity characteristics profile, in this example, for a "job-matching" application. Generally, a single competency is calculated by combining a set of occupation prior probabilities with context variable level probabilities, to make new prior probabilities for the next context variable as follows:

• Initial profile based on classification

o Firstly, the user selects a Standard Occupational Classification (SOC) code to the chosen level of specificity within the SOC hierarchy. A check is made that there are sufficient job profiles (typically sixteen) in the database in the selected class to allow a meaningful calculation to be made; if not, the parent class in the hierarchy is used, and so on until a sufficient number of job profiles are available,

o For a given competency the matching job profiles are split into four competency importance levels that describe how relevant ("Not relevant", "Partially relevant",

"Relevant" or "Essential") each competency is to the job profile,

o In the example shown, for the selected occupation of SOC code 17: "Architecture & Engineering", the matching jobs determined from database are broken down by number and proportion of jobs into the competency importance levels for each competency - in the example, the breakdown for Competency 1.1 : "Deciding and Initiating Action" is shown. o An occupation prior probability P(A) is calculated for each competency importance level from the number of job profiles within each of these competency importance levels, divided by the total number of jobs in the chosen SOC.

• Modified profile incorporating a first context variable

o To find the importance of a competency importance level given a selected context variable level, the number of job profiles in each context variable level is divided by the total number of jobs in that competency importance level, to give the conditional distribution P(B|A)

o These importances are then multiplied by the occupation prior probability P(A) to obtain the conditional prior probability distribution P(B|A)P(A).

o When the first context variable level is selected by the user the system divides the conditional distribution prior probability P(B|A)P(A) for the selected context variable level, by the total probability of said context variable level across all competency importance levels for this competency P(B), to obtain the new occupation prior probability P(A|B),

Bayes' theorem:

o In the example shown, First context variable (B): Organisation size"

• Modified profile incorporating a second context variable

o The values of P(A|B) are suitably rounded, and become the occupation prior probability P(A) for the second context variable, and the process repeats itself for said context variable and the following context variables, until the user has selected a context variable level for all context variables.

o That is, the posterior probabilities from the selected, first context variable level, now become the prior probabilities for the second context variable level, and the process continues as previously. The process iterates until the competencies given all the context variable levels have been evaluated.

occupation prior probability P(A|C), where from Bayes' theorem:

o In the example shown, Second context variable (C): "Education^*

If the value of P(A|B) is zero, the process halts to avoid errors propagating.

Figure 15 shows in tabular form the evolution of characterisation prior probabilities with successive iterations 1500. For the selected Occupation level, the initial occupation profile (defined by the probabilities at each of the four competency levels for each of the competencies) is modified in successive rows descending the table as successive context variables are taken into account.

Figure 16 shows in graphical form the evolution of characterisation prior probabilities with successive iterations. The effect of taking successive context variables into account can be seen on the four competency importance levels, essential 1600, relevant 1602, less relevant 1604 and not relevant 1606, evolving from left to right. Figure 17 shows an example of a single competency histogram. For the competency "Deciding and Initiating Action", the four competency importance levels are shown for both the initial predicted occupation profile 1700 and the final profile 1702 (as modified by the taking into account context variables), indicating that the effect of context variables may be considerable.

Figure 18 shows an example of a full final entity characteristic (competency) profile, displayed as a series of histograms 1800 with corresponding key. For each of the separate competencies 1802 the final calculated importance probability is shown according to the four levels ("not relevant", "less relevant", "relevant" and "essential").

In some embodiments, the user is presented with information about the probabilities of each of the four outcome categories together with a statement of the confidence with which the recommendation is made. Computation of confidence may be based, for example, on the variance of the four estimated probabilities and the number of jobs in the sample that these are based on.

Figures 19(a) and (b) show examples of an entity characteristics profile summary page 1900 - in the example shown, for a job matching application, this is a "Competency importance rating: Confirmation page". This page is merely to display the summary results (arranged in order of competency importance) for confirmation purposes. The user has the option to return to a previous screen to make amendments, to save the job profile and/or print the report 1902 - or alternatively to invite comment on the profile from other parties ("raters") 1904.

Figures 20 to 22 show examples of an interactive profile modification facility. Figure 20 shows a Rate competency importance page. Initially, only the system prediction is displayed on a stacked linear layout. Sliders are provided to allow the user to change the competency importance by moving the slider 2000 (which is both 'clickable' and 'slideable'). The system prediction remains unchanged throughout - as the user moves the slider away from the system prediction both markers will appear for that competency row.

Figure 21 shows an example of a hover-over / pop-up panel 2102, triggered by the user placing the cursor over and/or clicking a competency icon 2100, revealing additional information about the particular competency and explaining how the competency distribution resulted in the determined competency importance.

Figures 22(a), (b) and (c) show the Rate competency importance page with adjustment - the user having moved the slider 2000. Now both the slider original position (before editing) is shown 2000 and the users edited position 2200. A warning message 2202 shows that appears if the user, by his editing, breaks the banding rules placed on the system. The banding rule places a certain threshold on the number of sliders in any one competency importance level column.

Figures 23 to 26 show an external profile feedback facility. Feedback on the calculated profile may be sought from other parties, termed 'reviewers' or 'raters'. Figures 23(a) and (b) show a feedback progress screen, also termed a "Rater status" screen for tracking 'rater' progress - sortable by Name, job and status. Raters 1 16 can be contacted via email 2300 by the user 104. A (suggested) default email 2304 is provided (optionally editable via a pop-up text box) with the name 2300 of the user 104 and containing login details, brief instructions about rating, a link to the site to register (if a new user) or log-in (if an existing user) and a link to the job profile requiring their input sent to them. A copy of the rater invitation email is sent to the user (job creator). The user is prevented from sending invitations and/or reminders to raters that have already responded. Figure 24 shows the rater landing page, or the screen that will be seen by the 'raters' once they have clicked on the link within the email 2204. By clicking on the name of the job 2300, a new window with additional information will be brought up.

This screen shows the system profile prediction and allows the rater to make their own submission. If the Creator has completed the job (and closed all ratings) before the new rater gets the chance to submit their rating, the table on this page reflects that the job is closed - so there is no need to continue.

In some embodiments, the Rater is able to view more information about the job they have been invited to rate, if appropriate - such that, for example, they have enough information to provide a useful rating.

Figure 25(a) and (b) show the rater overview screen. In addition to the original slider position (before editing) 2000 and the user edited position 2200, rater edited slider positions 2500 are marked. On this page the user can observe the suggestions of the raters and has the option 2502 of selecting or deselecting which ratings are included on the display.

The various different rater responses can be displayed for example using different icons 2503, similar icons, or identical icons 2504 (as illustrated in Figure 25a). If identical icons are used for all raters, a further distinguisher may be displayed with the icon, such as a rater identification number. Optionally, raters may be grouped together in a single icon with a number indicating the number of raters associated with the icon 2506 (as illustrated in Figure 25b). The user may be given the option of which icon he would like his rating to be displayed with. A warning will appear if the rater, by his editing, breaks the banding rules placed on the system. The banding rule places a certain threshold on the number of sliders in any one competency importance level column. All predictions - including the banding rules - are displayed; banding ruling warnings are displayed only once the user (Creator or rater) starts moving their slider. Generally, the system prediction remains unchanged throughout; as a user or rater moves a slider away from the system prediction, both markers will appear for that competency row. If a user or rater makes a change that impact on the banding rules, that column will be highlighted with a warning message.

For the raters themselves, their view of the profile is similar to that first seen by the user, save they are only offered the option to submit a rating and not to invite other raters. On submission of a rating, an email is generated and sent to the user (job creator) confirming their submission. The email contains a link to the Rater status page. Once a rater submits a rating they are returned to the main system page.

Figure 26 shows the final profile review summary page. This is the page (only) the user 1 10 will see once all ratings have been received (the screen content builds as more ratings are submitted), and the user 1 10 has decided the importance rating for each competency (whilst subject to the banding rules). On this page the user has the option of saving and locking down the job profile 2500 or editing it further 2502.

Figure 27 shows the completion status page, where a user can select to download the report as a pdf, and also has the option of sending the report to others, via email. Information visible on this page includes: o Credit details

o Number of job scores available

o Number of enhanced reports available Checks are done to ensure that enough credit is available for the functions requested - for example, when adding an individual (to invite for assessment), downloading an 'enhanced' report etc. A copy of the email invite to individuals is sent to the Job creator / inviter.

Figure 28(a) and (b) shows the job library page, where the user can observe the current status (Open, closed, Completed, In progress, Not started) of all jobs. From this page the user can link to job resources that are available.

A job is 'in progress' for example when the user (Creator) has not completed the entire process or during the period when the job creator has finalised the competency selection and invited raters to comment.

In some embodiments, an option is provided for Job owners (Creators) to not only print the job profile, but also to view in read-only form the rater responses for each of the jobs they have created.

A 'Create new job' function is provided if the user does not find the job they require in their job library,

In summary, the user of the system described can be simply used to:

• Create a new job profile

• Optionally, invite others ('raters') to provide their own rating of the created profile

• Review rater feedback, adjust the profile and save it

· Invite individuals to take an online assessment to produce a match score against the profile

• Optionally, obtain a more detailed 'extended report on however many of these individuals as required.

In some embodiments, the system - by ensuring they start with a good set of competency requirements - enables person-job matching to be carried out by relatively unsophisticated users. Having decided on the competency requirements, the system will select a set of assessments designed to optimally measure the competencies and provide the means of online administration of these to specified candidates. The system then uses a matching algorithm to compare each person's score profile to the competency requirements and provides a report to the user on the relative goodness of fit of candidates to the position.

The selection of assessments may be optimised to measure competencies in dependence on the previously determined competency requirements. A multitude of tests is available to measure a wide range of competencies, with specific tests measuring specific competencies or possibly groups of competencies. It may not be necessary to subject candidates to broad, exhaustive assessment, but only a subset that measures competencies of particular interest. A competency weighting may be associated with each competency, according to importance class. For example, only competencies above a predefined threshold may be used to determine the selection of assessments. All assessments required for the selected competencies may be selected.

Further, an assessment that is appropriate to an essential competency (that is, a competency that is above the pre-defined threshold) may also be used to test a non-essential competency. Thus it is possible to evaluate competencies that are otherwise not considered sufficiently important to merit dedicated assessment. Alternatively, a combination of assessments that covers the greatest exposure in the weighted competencies for a given limited number of assessments may be determined. In this manner broad, exhaustive assessment may be avoided, therefore streamlining effort and optimising outcome. For evaluating the assessment data derived from the assessments, a weighting scheme may further tailor the result to the user's needs. As described above, the assessments provide information regarding the competencies that are essential, but they can additionally provide information about further (nonessential) competencies. By using the assessment data that is available in any event to evaluate further competencies, the information extracted may be increased at no additional assessment effort. A score that relates to competencies that are important but not essential can be combined with the scores that relate to essential competencies by weighting the scores according to the associated competency weighting. By combining scores of different relevance (as described above) - or otherwise - in the preferred embodiments a simple assessment (in the form of a single mark, score or rank) is provided for each candidate. This facilitates interpreting the score results, as the result provided to the user could be as simple as a single 'job suitability' score for each candidate.

Figure 30 illustrates an example of the selection of assessment instruments (typically comprising a series of individual assessments or assessment methods/techniques) and interpretation of scores as described above. A particular competency 3000 of interest is determined based on whether it has an associated competency weighting above a particular threshold (here: weight 4 out of 4, but in other embodiments weight 3 or 4 out of 4). For the particular competency 3000 of interest an assessment instrument 3002 is generated that combines a number of individual assessment methods 3004 (here OPQ1 , OPQ3 and OPQ4 for a first assessment instrument, and OPQ7, OPQ10 and "Verbal skills" for a second assessment instrument) that are known to be particularly relevant for determining that particular competency 3000.

In assessing the scores from the assessment methods once the assessments have been undertaken by the relevant candidates (and again with reference to Figure 30), it is possible that one assessment method 3004 is of greater relevance than another, in which case the scores from the individual assessment methods are weighted. In any event, the scores from the individual assessment methods are combined as appropriate to form a competency score. The calculated competency scores are weighted using weights 3006 and combined in order to present a single score for the particular candidate. With such a system, informed selection of candidates is possible for a user without specialist knowledge about each type of assessment, what competencies it evaluates, and how to interpret the results. Instead, the user may rely on the system to optimise the assessment process to the user's requirements, and evaluate the results in accordance with the user's interests.

In the example shown in Figure 30, some of the assessment methods (here: OPQ1 and OPQ10) are appropriate for assessing a further competency 3008 that was not specifically selected for testing (here: weight <4). In a variant, the individual assessment methods may be used to determine a competency score for the further competency 3008. As for the other competencies one assessment method may be of greater relevance than another, in which case the scores from the individual assessment methods are weighted at 3010 and combined as appropriate.

The system may be integrated in a platform that enables further functionality, for example administration of applicants; submission of invitations to candidates to undertake an assessment; administration of assessment results; collecting, scoring, evaluating and/or reporting of assessment results; presentation of assessment results; and/or other functions relating to a characterisation profile and potential candidate entities. Further Features

The following relates to commentary on the system described and further features which may be included:

• In the system, three types of competency importance classifications are distinguished:

^■ Predicted importance: what the Bayesian estimation process has predicted the importance to be.

■ Rated importance: what a rater of the job has rated the importance of a competency to be

(rater could be the "job owner" or an invited rater).

^■ Selected/Final importance: The final importance selected by the job owner for each competency, having taken into account the predicted importance and optionally all instances of "rated importance" by invited raters.

· The database stores all types of importance profiles for a job: Predicted importance; Rated importance (could be more than one if several raters); Selected/Final importance.

• A Multi-rater facility allows other users, specified by the original client, to moderate the competency importance levels output by the system, by rating them. The system would then use the multi-rater input (ratings) as data for the second iteration of the Bayesian estimation of the profile (starting where job database estimation left off).

• The system is "self-learning", with the created competency profiles being fed back into the database and used in the estimation of new profiles.

• The classification system is based around the standard occupational classification or SOC, but the program has potential to move beyond this framework.

· Use of a Universal Competency Framework (UCF) with, for example, 1 12 competencies or alternatively, for example, 93 competencies at the component level. The client competency models can be mapped to the UCF, so the profiling process can be used with a client model/language which allows for great flexibility.

• Ways in which competency profiles today are determined include conducting analysis of the job in question, for example:

• Directly measuring the competency profile of the job using Card Sort. This can be quick but is difficult to standardise and validate.

• Deriving the competency profile from a detailed task analysis of the job using a Work Profiling System (WPS). However, this is time consuming, needs experts but is rigorous, standardised and provides good validity.

• Consulting a database of jobs/occupations allows for mapping of, for example, the 0*Net job database content to the UCF and the generation of profiles from that. This is simple but relies on quality of job analysis data in 0*NET, and relevance of chosen category to actual job - and ignores the strategic use of competencies.

· Most job databases, like 0*NET, do not have any job-level data but rely on generic data.

• Based on thousands of detailed job analyses using the Work Profiling System (WPS). Each job analysis can take days to complete but provides the most robust data on which to base competency requirements.

· WPS has been mapped to the UCF and so component and dimensions level profiles can be generated for any job analysed with WPS. • In order to account for every job being different, over 7000 jobs have been mapped to the SOC codes used in 0*NET. This means it is possible to look at the similarities of the competency profiles generated for individual jobs within each job family as defined by SOC.

• The importance of a competency is partly a function of the job family and partly a function of organizational context.

• It is possible bring together the structure (SOC) used for making 0*NET practical with the richness of the data that has been collected over the years using WPS to underpin the process of competency profiling.

• Rather than giving the user a 'blank' sheet and then asking the user to pick what the user thinks are relevant competencies, the user can be provided with a lot of information about what competencies are typically important or not important in his/her jobs/roles.

• If that is the starting point, then the users are more likely to end up with a profile of requirements that accurately reflects the job: you can build on the foundation of many thousands of job analyses rather than starting from scratch each time.

· Each new profile is derived from hundreds of previous ones.

• Map client competency models to UCF, so you can use the system profiling process with client model/language.

• Link to products: products are mapped to competencies, so it is possible to immediately suggest which products to use to assess for the selected competency model.

· Estimate probability of WPS (job analysis) task items: Can be used as first-draft job description, or as a first-cut of which items to respond to in a job analysis survey.

• Move beyond the OC as the framework users interact with (can still use it behind the scenes).

• Coverage gaps may be addressed by drawing on 0*Net data

• In those cases where there is no or insufficient data in the WPS database, competency requirements can therefore be generated from equations that relate UCF (or other competency model) to 0*NET variables.

• Bayesian methods may also be used to combine data from various sources - WPS job data;

0*NET job data; date from users of the system - and suitably weight the value of the different sources of data.

· Users may also use the system to create and rate jobs, and combine jobs created by these different users (for example users from the same organisation) into an integrated set of jobs. In other words, uber jobs or requirements may be made out of smaller completed sessions.

• (Quick) Registration page

• Pre-populated with the details input by the job creator - the (new) rater just has to validate them

• User inputs more detail, creates a password and submits their registration

• The system generates a verification email that contains a link to the registration page (so they can log in)

• Rater registration page

· On receipt of the verification email the user clicks to go back to the registration page - and complete the registration process - AND upon registration be directed straight to the Rater landing page.

• Buying page / basket

• Allows buying of Job profiles, job scores, enhanced reports.

· All three items listed as available to purchase - User adds whatever quantity of each they require. • Checkout page - allows a user to us existing process for checking-out once the items have been added to the basket; payment on account or via credit card.

• A confirmation email is sent to the Job creator once the credits have been purchased (as per existing process).

• One example of how to determine whether a context variable is deemed "relevant", is to make use of Kendall's Tau-C statistic, which indicates how strong the relationship is between two categorical variables (such as a context variable with, for example, 5 levels versus a competency variable with 4 importance levels). The Kendall's Tau-C value is between 0 and 1.

The following steps are taken:

• The Kendall's Tau-C statistic is calculated for each combination of SOC occupation X Context Variable X UCF competency. For example, this statistic may be calculates for the relationship between the context variable "Level of Education" and competency "Deciding and Initiating Action" for the SOC occupation "Marketing Managers". For the same SOC occupation, the Kendall Tau-C metric is continued to be calculated for all combinations of Context Variables X Competencies (21 Context variables X 20 Competencies).

• These analyses are repeated for all SOC occupations at all four SOC levels. · Some cut-points are then used to determine how large the Kendall's Tau-C value needs to be for a relationship to be considered important and how many competencies a context variable needs to have a meaningful association with for the context variable to be included, for example:

^■ Minimum Kendall Tau-C cut-point: 0.1

■ "Substantial effect" Kendall Tau-C cut-point: 0.2

^■ Minimum number of competencies affected with at least minimum Kendall Tau-C value: 4

• Decision rule on which context variables "matter" for one SOC occupation. Variable included if it

■ affects at least 4 competencies with a Kendall value more than 0.1

^■ affects ANY competency with a Kendall value more than 0.2

In this embodiment, these current analyses are only conducted on the UCF20 competency model; the same or similar analyses may also be run on other competency models. In alternative embodiments, context variable relevance may be determined using different cut-points - or different methods altogether.

These procedures may be re-run periodically as more data is added to the job database, as the determination of which context variables are most relevant will become increasingly accurate as data is accumulated.

Figure 29 shows an illustration of how the relevance of a context variable is determined, in the form of the calculation for one SOC occupation. Shaded boxes 2900 indicate a likely relationship between competency and context variable (i.e., that the importance of competency varies with the levels of the context variable).

• Occupation-dependent context questions: The set of job context questions that are asked of the user (and that form part of the input into the Bayesian estimation procedure) are dependent on the occupation code selected by the user for the job. A piece of analysis has been conducted to determine, for each occupation code (at each of the four levels), which set of WPS-based context questions are most likely to be informative about the predicted importance of competencies. This ensures that the user is less likely to have to respond to context questions that will not make a material difference to the estimated competency profile.

• Competency task statements: For each of the competencies, a short matched "task statement" has been written that illustrates the task/activity context where the competency will tend to be required. For example, for the competency "Deciding and Initiating Action", the matched task statement is "The job requires making decisions and taking responsibility for them. Typically it will involve making tough choices based on careful consideration of risks, taking responsibility for such decisions, or taking action on one's own initiative". The intention of these task statements is to illustrate the competencies in task/activity terms, rather than in terms of personal requirements. If users think of the competency constructs in terms of what activities are actually required in the job rather than which personal qualities they would like to see in job holders, they are more likely to accurately judge the importance of the competency constructs.

• Task/activity oriented rating scale: This is related to the previous point. The rating scale on which the users are asked to indicate how important each competency is to the job is task/activity oriented. This contrasts with the personal requirements oriented rating scale ("Essential", "Relevant", "Less Relevant", "Not Relevant") that tends to be used. The final competency profile can still be phrased in personal requirements terms using logic such as "because this competency is a defining part of the job, it is essential that job holders are good at this competency". The task/activity rating scale is as follows:

o "Not part of the job": These activities are not part of the job. If a job holder performs these activities poorly, there is no risk that the overall objectives of the job will not be reached."

o "A minor part of the job": These activities are not important to the essence of the job. The job role would remain the same even if these activities were excluded. If a job holder performs these activities poorly, there is a small risk that the overall objectives of the job will not be reached."

o "A significant part of the job": These activities are important to the essence of the job. If several of these activities were excluded from the job role, this would be a completely different job. If a job holder performs these activities poorly, these is a substantial risk that the overall objectives of the job will not be reached."

o "A defining part of the job": These activities are key to the essence of the job. If any of these activities were excluded from the job role, this would be a completely different job.

If a job holder performs these activities poorly, it is almost certain that the overall objectives of the job will not be reached."

Alternative embodiments may implement some or all of the following features:

• Link the predicted competency profile to a large catalogue of assessments, which in turn is linked to a meta-analysis database of validation evidence. When combined, this provides the functionality to predict the competency requirements are for a job, to recommend which assessments would be optimal to assess individuals in terms of their potential to match the requirements of the job, and to predict the level of validity to expect if the recommended assessments were used to assess for the job.

• Predict probability of (WPS) task items: Using the same Bayesian estimation process used for estimating competencies (with the same inputs: occupation code and context variables), instead the probability of WPS task items are estimated (examples of such items are "Scanning written information", "Creating appealing presentations", "Cataloguing material". In essence, competencies are swapped out for WPS task items as the dependent variable of the estimation. This is possible because the jobs in our WPS job database contain information on both competency importance scores as well as ratings of importance of hundreds of specific tasks. There are several possible applications of such estimated task importances:

o Create a first-draft task-based job description: The competency estimation process helps a user specify the competencies required of a person doing a job. Estimating task importances would help the user list the specific, detailed activities the job holder would need to perform.

o Create a first-draft of a job analysis survey: The task items predicted as most likely to be important could be used to construct a job analysis survey where job stakeholders are asked to directly rate the importance of each item for a job. This contrasts with scenarios where such stakeholders would have to rate a larger volume of tasks where many such tasks are unlikely to be relevant for the job.

o Shortened WPS process: The process for conducting a full WPS job analysis process requires an initial phase where the set of WPS task items to be responded to are narrowed down by, first, selecting the appropriate version of WPS for administration, and second, conducting a card-sort to determine which questions within the selected version to respond to. Drawing on a set of items with estimated importances could bypass these two steps.

• Run estimation procedure for client competency models: Functionality can be designed where a client competency model is mapped to the models in the estimation system, and predictions can be generated and reported directly into the language of the client competency model.

The following table shows examples of "user stories", illustrating potential uses of (and further information relating to) the described system, to be understood in the format "As a ... I want ... , so that .. ."

New job Variable number of User to have to answer only 1 don't waste time profile 'context variable relevant context variable answering questions creation questions' questions based on the job that are irrelevant that 1 have selected from the

static list

New job Descriptions of Product to show users some additional users are more likely to profile context questions Owner information about the context respond consistently creation variable questions - to put the

context variable questions into

context!

New job Display predicted User to see the predicted 1 am able to review them profile competency competency importance

creation importance ratings (UCF20) for my job

[3] once 1 have entered my

responses to the context

variables

New job Display certainty of Product to be able to tell users how users can understand profile predicted Owner sure we are in the competency how comfortable they creation importance ratings importance prediction that we should feel in making make changes to the

predictions

New job Save changes to User to save the importance ratings 1 either accept all profile predicted once 1 have reviewed them predictions (having creation importance ratings reviewed them) or save [4] the changes that 1 have made to predictions

Additional Review multiple User to be able to see other job 1 can make final

Ratings ratings rater views on the competency decisions on the profile

[5] profile alongside mine that will be used to assess candidates

Report Printable/saveable User to be able to print or save a 1 have a physical record generatio job document that shows the job of the job profile n profile/description profile 1 have created

document (including some standard

area's for me to enter further

information)

Candidate assessment User the job profiling tool to 1 do not have to be an assessme recommendations recommend the assessments expert to use the tool nt that my candidates should

complete based on the final

job profile

Candidate Invite candidates User be able to enter candidate my candidates are assessme names and email addresses assessed

nt

[6] New job User job library User to be able to browse all jobs 1 1 can easily access profile have previously created on the previously created jobs, creation system and either edit the job,

[7] invite candidates to assess against them, or clone them as a basis for a new job.

New job Organisation job User to be able to access a list of all 1 can select a relevant profile library jobs created in my existing job for editing, creation organisation by myself or other cloning, or assessing users from my organisation candidates, or

inspect/build relationships between jobs in my organisation

New job Associate job with User to make sure a new job is we can keep track of profile organisation automatically associated with which jobs from the creation the organisation the job owner same organisation that belongs to have been created on the system

Overall Create new Central to be able to register new so that the job owner system organisation Administ organisations on the system users can be associated rator with the organisation they belong to

Overall Associate user with Central to be able to associate a job the organisation context system an organisation Administ owner user with an of a user is available rator organisation when the users creates new jobs

Design Extensible Product the backend architecture of the so that 1 can easily architecture Owner application to be flexible, extend the functionality modular and extensible the profiling system exposes

Data Access to job Product easy direct query access to all 1 can pull out job

Extract database Owner aspects of the SQL database information to enhance

[8] where job and usage data is and improve the job being stored database underpinning the application

Registrati Register new User to enable the system to job profiling activities on external user register a new external user on can be performed on

[9] the profiling system behalf of the external user

Data Get list of Central to be able to get a list of all 1 can review and

Extract registered external Administ external users from my manage the users that users rator integrator user account that have been registered have been registered on the

profiling system Overall Get list of Central to be able to get a list of all so that I can review and system registered external Administ external users from a specified manage the registered users from an rator organisation that have been external users organisation registered on the profiling

system with my integrator user

account

Data Get all Central to get a list of all organisations so that I can review and Extract organisations Administ that have been registered to manage the

registered to rator the integrator account on the organisations my users integrator account profiling system are located within and are using on the system

New job Get list of jobs in User to get a list of all jobs I can check whether the profile organisation registered to my organisation job I am profiling has creation been profiled before by someone else, or I can find a job that is similar to the job I want to profile, and I can use it as a template

New job Get jobs that are User to be able to get a list of jobs I can review the job profile hierarchical child that are hierarchical child jobs content and competency creation jobs of a specified of a specified job (i.e., that requirements of jobs at job report to the specified job) different reporting lines and levels in relation to each other

New job Set parent job User to be able to set the parent job I can manage the profile of a specified job hierarchical creation relationships between jobs

Data Access to ALL job Central to be able to access details of Assist with customer Extract profiles Administ ALL Job profiles created in the enquiries etc....

rator system

Data Ability to extract job Central to be able to extract all Job I can use the information Extract profiles from the Administ profile details at any time from to update the database system rator the system of job profile information

- so that it is kept constantly up to date

Additional option to exclude User to be able to exclude certain I can focus on specific

Ratings some ratings raters importance rating scores importance ratings that

[10] from the summary page view are not skewed by any one individual rater

Additional New rater viewing Invited contributors will not have I can add my own Ratings job profile being rater access to Screen 1 and Screen ratings to each of the created 2, but Job Name and Job competencies

Description to be displayed in

all Screens (except for Screen

1 ); and to view the summary details with predicted

competencies importances

provided by the system but

NOT any ratings submitted by

other invited raters

Additional Submit competency Invited Add my competency The role profile created

Ratings ratings rater importance ratings to the Job can include feedback

[1 1] profile being created from a number of

different people involved in the role

Overall Prevention of Product to be able to prevent the Users cannot abuse the system amendment of Owner amendment of existing Job system, simply

existing job profiles profiles amending existing profiles rather than creating new ones each time

Data Access to usage Product to have access to usage I can track the success Extract details and Owner statistics (number of profiles, of the solution

statistics reports, assessed candidates

etc )

Overall Future proofing to Product to have the solution future When the third party / system include third party Owner proofed to incorporate third UCF mapping has been tests party tests and assessments completed, more tests and assessments can be offered as part of the assessment process

Candidate Use Candidate Product to use the existing candidate there is a consistent assessme management Owner management functionality solution offered to nt process already available to manage customers

the assessment process

New job No effect from Product to have updates made to the Role profiles that have profile underlying Owner underlying database, but not been created remain creation database updates have these updates affecting valid and can be

[12] any Role profiles that have referenced by users at been created and Saved. any time after their creation (even if the profile would be likely to change if it were recreated, due to the updates to the database). Registrati Granting access to Central Review the users' application I can make a decision on Central Administ and access information about as to whether to grant rator the individual them access to Central

Registrati Approval of Central Central to approve or reject a user's I can give access or on access Administ application and send deny them access and rator notification email notify them

automatically of the decision and impact thereof

Candidate Comparison of User to be able to compare I can make a direct assessme candidates at candidates that have been comparison of nt different times assessed against the same candidates over time

[13] role at very different times - (OR include additional

AND also be able to candidates at any time distinguish between different during the assessment groups of candidates that have process)

been assessed against the

same role profile at different

times.

Additional Pre-identification of User to be able to invite raters - they can rate the job

Ratings invited raters invite them using an email with profile I am creating

[14] a unique URL that will take

them straight to the registration

page then contributor's landing

page

New job Option to choose User to not have to pick from all 4 of the process is a simple profile from drop down the drop down lists - put as a as possible. One creation boxes minimum I must select from selection MUST be [15] box 1 made to ensure that the a high level competency selection can be made

Overall Class leading ease User once I have selected the Job the Job Profiler is easy system of use - for ALL Profiler option to view a to access and simple to screens simple, easy to understand, use.

uncluttered landing screen

New job View competency User to view ALL the competencies the competency review profile summary on one page - All 20 of them. and amendment creation The predetermined importance process is as simple as rating for each one of the possible

competencies will be pre-filled

and the user must have the

option to click on any one of

the other imp ratings as

required. SCREEN could Additional Invitations to raters User to have NO restriction on the I can invite raters

Ratings number of people I can invite

[16] to rate the role profile I have

created, have fields available

to input: - Rater name, email

address, relationship to me (by

choosing from a drop down list

of options),

Additional Slider functionality User Slider functionality for I can finalise the job Ratings on screen 3 Creator/Rater to rate profile immediately competencies. It should also

display predictions

Candidate Generating PJM User to have the option to generate I can see more detail for assessme reports a PJM report for any number any one of the nt of the candidates assessed candidates at any time

[17]

General Ease of User to know exactly where I am in I am clear as to how screen understanding the the Job Profile Creation many steps there are in navigation creation process process - perhaps using a the process - and how and screen flow STEP process, highlighting at many of them I have the top of the web page what completed

steps there are and which

have been completed.

New job Making User to be able to go back and I can correct any errors I profile amendments to role amend details I have believe I have made creation profiles submitted when creating a new AFTER the event

[18] role profile - BEFORE I have

saved the role profile

Candidate Track assessment User to be able to track progress of I can see what's going assessme progress candidates being assessed on with the assessment nt

Candidate Chasing candidates User to be able to chase candidates I can control the assessme (if required) that are involved in assessment project nt the assessment Individual Candidate User Create a job profile and assess I don't have to ask

Store assessment against against that role with precandidates to sit the

[19] multiple roles existing test and assessment same assessments results - i.e. Be able to assess numerous times.

ONE candidate against a

number of different roles

(providing the minimum

assessment requirements are

met)

Overall Data Protection Product The Data Protection Notice for I can keep the system system Notice content Owner ALL users (Role creators and up to date with the latest role raters) to include information captured permissions that allow for the

data submitted to be used to

update the underlying

database.

New job Cloning options User when cloning a job profile to be it is easy for me to profile able to view ALL the inputs create a new role that is creation previously made for the job very similar to an

profile AND be able to make existing role on the amends to any of the inputs I system

wish to. In addition, I would

want to be able to change the

name of the Job profile (this

information would

Candidate Report options User The 'purpose' of an activity it is as uncomplicated as assessme available should be determined once a possible for untrained nt User has confirmed that they users to get what they wish to assess against a job want from the system. profile they have created.

Additional Inclusion of a Invited to have the option to include I can let the job creator

Ratings comments field for rater comments when I add my know anything else that

[20] raters ratings for a certain job profile I feel is relevant to the job profiles being created.

Additional Multiple ratings Invited to be able to complete ratings the rating process for Ratings rater for all the profiles I have been multiple profiles is as asked to rate in one go - i.e. painless as possible. not have to use the link from

each of the separate emails I

have been sent

Additional Landing page for Invited to be directed from the email I I can submit the best Ratings invited raters rater receive inviting me to rate to a rating possible.

landing page that offers me

instructions and information on what is required of me / things

to consider etc..

Additional Resend email User to have the option to RESEND any raters that say that Ratings functionality any rater invites have not received the email with the unique link can receive it and then go on and rate the job profile

Candidate Request additional User to have the option to order I can view a more assessme reports additional (more detailed) detailed report ONLY for nt reports by using the links on those individuals that I the 'Completion Status' page need to / want to....

Candidate Select purpose for User once I have chosen OR Once I have chosen the assessme candidate created a role profile to choose reason for the nt assessment an option to assess candidates assessment it pre- and then be taken to a screen populates the screen where I select the purpose of where I complete all the the assessment (from individuals' details Selection OR Development)

New job Making User to have the ability to make for example, I can stop profile amendments to amendments to competency half way through creation competency estimation input parameters - creation of a profile and [21] estimation input BEFORE user requests the come back to it at a later parameters - benchmark (in Screen 3) time to complete and BEFORE user SAVE the profile requests the

benchmark (in

Screen 3)

New job Banding rules for Product to ensure that there are limits we don't have situations profile competencies Owner to how many competencies where too many creation ensure that there that can be classified into each competencies are

are limits to how importance band classified into the higher many importance bands competencies can

be classified into

each importance

band

Implemen Developer Impleme it could be integrated with the

t WSI OD nt WSI front end screens at a later

back-end OD date

back- end for

JPT System

Testing

UAT

Support

New job Saving inputs when User To be able to submit should 1 move profile creating or cloning responses to the additional away from that page creation a new job (context variable) questions and come back to it, the and for those responses to be responses 1 gave saved in the database. This previously are still means that 1 can go between available for me to view screens and not lose any of / amend.

the responses 1 have input to

either of the screens. In

addition, if 1 am cloning a job

BUT 1 choose to change the

original foundation job (new

SOC code) the system will

return the results from the

'supplementary' questions that

are consistent across the two

different roles being

considered. Any previously

submitted responses that are

not relevant to the new job

being created (from the

cloned job) will be deleted so

that they have no impact on

the new job.

New job There is a User to be able to input the level of The 'system' can make profile requirement to ask the job 1 am creating. This the correct selection of creation the Job creator to input will be by choosing the which Verify level (etc.) explicitly state what appropriate option from a drop to include as part of the level of job they are down list. assessment suite to looking to define / create the job score. profile.

Overall System needs to be User to be able to use the system in The service can be used system localised rather a language and more technical with no issues outside of than just translated terms that are familiar to me. the UK.

(e.g. US$ referenced and not

UK£). There is also the issue

of payment currency to

address, but this can be picked

up at a later date. Overall System Admins Administ To be able to help users by Customers that make system (working on rator having access to everything mistakes/errors when

Helpdesk) need to they can do on their system using the system have a have access to and make changes even when support mechanism that make changes to they are not able to. they can use - that can any user inputs as actually help them. required.

New job Original rater User To have my first rating In future analysis of job profile (Creator) submission saved, just as it is creator behaviour can creation submission is for all other raters. Any be analysed - how

saved, not just changes 1 make when finally frequently do raters overwritten when confirming the job profile WILL change their submission the FINAL profile is NOT overwrite my original based on the responses confirmed. rating submission, but will be of other raters for

saved separately. example

Data Offers the user the User to be able to extract details of existing clients with their Extract opportunity to carry the job 1 have just created own systems and are out two distinct using a format that can be able to use the Job tasks: Profile a used in other systems profiler and then take specific job AND (primarily xml format that can the results and use them use the job profile be uploaded). on their own system. created to assess

candidates against

(and generate a job

match score).

There is a need to

be able to extract

the full details of

the job profile

created from the

Central system in

order that it could

be used elsewhere

- notably with the

PJM functionality

directly.

New job Need to input User To be able to input the The information profile additional following information: captured can be creation information about Organisation, Department, included in the Job the job being Location matching competency created - so that report

the information is

available in the Job

profile report New job When the User to be able to see clearly which 1 can create a focused profile prediction is competencies 1 should focus job profile as efficiently creation returned back to my efforts on based on the as possible,

the Job creator it predictions that are least concentrating my efforts should be possible certain as highlighted to me. on those areas where to highlight those they are most required. competencies

where the system is

'less certain' about

the Importance

ranking prediction

we have made. All

details of the

prediction are

available from the

graphical display

that is accessed by

clicking on the

competency name,

but it should also

be possible to

highlight to users

where the

prediction is less

certain (perhaps

limited to 3

competencies we

are LEAST certain

about)

New job Require the options User to VERY easily navigate the 1 can save time with my profile offered in the 4 job list options to be able to job selection - improving creation drop down find the best match as simply the overall process for selection boxes to as possible. me.

be as easy to

understand and

navigate as

possible New job Need to be able to User to have the option of choosing 1 can save time with my profile offer the Creator from a list of jobs that are job selection - improving creation the option of chosen by users most often the overall process for choosing from a AS WELL AS being able to me.

summary list of choose from the drop down

'Most popular' job lists available to me.

titles - to make the

job name selection

process quicker

and more

straightforward

New job to have a Job

profile Hierarchy in the to have a Job Hierarchy in the

creation Job Profile page User Job Profile page

Notes on the above table are as follows:

[1] 1. Create a job table in database

2. Create a stored procedure to save jobs in database

3. Create a job controller with methods

4. Create a job view

[2] Once this selection has been made and the user clicks NEXT to navigate to the next page

(Context variable questions) the charge occurs and the user cannot make any change to the previous choices - only changes to the subsequent pages.

[3] 1. Call Estimation Engine

2. Populate Ul from Estimation Engine

3. Allow users to select competency importance

[4] Saving the ratings you have agreed will mean in effect that you have SAVED the profile as a

whole. NO changes possible after this point.

[5] Display ratings from multiple raters (and predictions)

[6] Using the existing 'Completion Status' page, need the following fields to complete / view:

Purpose of assessment (Selection OR development)

Email address

First name

Last name

Title

(optionally) Language (will determine what language the individual is first offered for assessment) Gender

(optionally) Product

Job Name

Status

PJM score

NEW (view only)

OPTION to order additional reports

[7] Task 1 is the tabular view of jobs with the action options available etc. Task 2 is a tree / org view of jobs crated (To be completed in future iteration - TBC)

[8] Need to be able to gain access to the underlying data in the system AND be able to update it - but this should all go through the existing protocols. No special access to the database is expected.

[9] Access to the Job profiler functionality is 'free' to all Central users, so this requirement is already available as part of the general Central functionality.

[10] view 2 - display ratings from multiple raters (+ predictions)

[1 1] viewl - creator - view summary; Lock the profile & proceed to job library (to print Job profile);

Invite raters; ; back button - 0.5 pd

view2 - rator - view summary; Lock the profile & Submit ratings (it will take them

to the landing page); back button - 0.5 pd

[12] Profile created remains fixed, so that there is a constant record of the Job profile created on the system. If users wish to take advantage of system updates they have to create a new profile which may or may not be different from one created with the same input parameters as before the update was carried out.

[13] No candidate grouping

Any review of candidates that have been assessed will be carried out using the filters available on the system (ie. Sort results by date etc - then individuals can then be compared. No other facility is required.

[14] All raters must register before taken to the landing page.

[15] 1. Create tables for occupations with foreign keys, etc.

2. Create a stored procedure to read the list of occupations

3. Add new methods to the job controller to retrieve the list of all occupations and save it into the cache

4. Add 4 new drop down lists to the job view to display occupations

5. On selection change update sub-occupations by querying the cache with LINQ

[16] Must include details of suggested number of raters etc. in the HELP text on the page - but have no system restrictions.

This is the Invite Rater screen story. Estimates updated to 5 pds.

Taskl - Sorting and Filtering

Task2 - Invitation Links

Task3 - pagination

Task4 - Add buttons for Invite & Resend for each rater and each job

Task5 - Customise email link

[17] no selection drop-down, just additional button on Completion status page

[18] Once a job profile is finally submitted, all screens become read only. If user decided to lock down the ratings and generate report, then the contributor who had not completed the ratings will not be able to do it and get error.

[19] Individual store

Assessments against roles always start with the role and then lead on to the assessment (rather than the other way round). Expected that when details are input in the 'invite individuals' section that there is recognition if the individual is already 'in the system' - the role new role profile can then be assessed against the results that have already been used.

For the same candidate is assessed for 2 jobs, the scores for job 1 will affect the scores for job 2

[20] Just one free form text field - OPTIONAL for raters to submit any comments

The application responds to the growing need for organisations to have real confidence in matching the right people to the right jobs, so that they perform to their highest potential. The system provides:

• a quick, easy and highly reliable means of measuring the fit, or match of an individual's competencies and potential, with the specific competency set required for a particular job.

• a robust way of creating a competency based job description for a specific job, and then assessing the degree of people's competency match with that job, providing an overall view of an individuals' fit to a specific role and can be used in both a selection and development context.

• a straightforward online step-by-step process, designed for use by line managers at any level in any organisation, not only by trained HR professionals.

• the job analysis and specification on over 7,000 jobs.

• Four main steps, each one completed online:

1. Define the job to be assessed for either recruitment or development by creating the job description with an easy-to-use, automated job description / competency builder.

2. The job description / competency builder is based on more than 7,000 jobs mapped to the SHL Universal Competency Framework.

3. You then have an option to share the draft job description with other stakeholders, inviting them for their input, which you can then consolidate into one overall job view, making changes to the final job description if you wish before locking it down.

4. Once the job description has been locked down, you then invite individuals to sit a couple of assessments, determined by the solution and tailored to best predict the competencies included in the job description. This results in a score and report showing how closely an individual fits the competency requirements specific to that job.

• There is also an optional, more detailed "plus' report, giving you a bespoke interview guide specifically designed to help you explore that individual's stronger or weaker areas of fit to the job.

It will be understood that the invention has been described above purely by way of example, and modifications of detail can be made within the scope of the invention. For example, in the case of requesting replacement machine parts, this could be achieved by the method as described earlier, mutatis mutandis. In such a method, the classification parameters of the present invention might for example be a description and/or function of the particular desired machine part, and the characterisation parameters might be properties of a potential replacement, with the requirements characterisation profile describing a potential replacement, constructed by posing suitably crafted questions to the person requesting the replacement machine parts based on the characterisation parameters.

Each feature disclosed in the description, and (where appropriate) the claims and drawings may be provided independently or in any appropriate combination.

Reference numerals appearing in the claims are by way of illustration only and shall have no limiting effect on the scope of the claims.

Claims

1. A method of determining a requirements characterisation profile for an entity, comprising the steps of:

receiving classification parameters defining a requirement for an entity; selecting, in dependence on the classification parameters, a set of entities from a database of previously assessed entities;

retrieving from the database characterisation parameters of the selected set of entities; and

constructing, in dependence on the characterisation parameters, a requirements characterisation profile for the entity.

2. A method according to claim 1 wherein the classification parameters allow an entity to be identified generally.

3. A method according to claim 1 or 2 wherein the classification parameters are in the form of a nested list of classes and sub-classes.

4. A method according to any preceding claim wherein only a subset of characteristic parameters is used to select the set of entities from the database.

5. A method according to claim 4 wherein only the most significant characteristic parameters are used, the most significant characterisation parameters comprising those deemed to be at least one of most relevant and least relevant.

6. A method according to any preceding claim wherein the characteristic parameters are competencies.

7. A method according to any preceding claim wherein the step of receiving the classification parameters comprises determining the classification parameters from answers to a series of questions concerning aspects of potential entities or candidate entities.

8. A method according to any preceding claim further comprising the step of receiving a set of context parameters or variables.

9. A method according to claim 8 wherein the context parameters or variables are determined in dependence on the classification of the entity.

10. A method according to claim 8 or 9 wherein the number of context parameters or variables used is in dependence on the classification of the entity.

1 1. A method according to any preceding claim further comprising the step of:

determining the probability of a randomly selected member of the set of selected entities having a characterisation parameter in an importance class.

12. A method according to claim 1 1 when dependent on any of claims 8 to 10 further comprising: determining modifiers with which to adjust the probabilities in dependence on the context parameters.

13. A method according to claim 1 1 or 12 further comprising determining the probability for each of a set of classification parameters for each of a set of importance classes.

14. A method according to any of claims 1 1 to 13 wherein the context parameters or variables used are those which are determined to have potentially greatest effect on the resultant probabilities.

15. A method according to claim 14 wherein determining whether a context parameter or variable has potentially the greatest effect on the resultant probabilities uses Kendall's Tau-C statistic.

16. A method according to any of claims 8 to 15 wherein context parameters or variables are determined from previously determined context parameters.

17. A method according to any of claims 8 to 16 wherein the context parameters or variables are determined from answers to a series of questions concerning aspects of potential entities or candidate entities.

18. A method according to any preceding claim further comprising assigning the classification parameters to an importance class.

19. A method according to claim 18 wherein the importance class is one of: essential, desirable, relevant, non-relevant.

20. A method according to claim 18 or 19 further comprising the step of:

determining the fraction of members of the selected set of entities with characterisation parameters in each of the importance classes.

21. A method according to claim 12 wherein the probability adjustment comprises a Bayesian probability calculation.

22. A method according to any preceding claim wherein the method further comprises determining a characterisation profile or model determined from the classification parameters.

23. A method according to any preceding claim wherein the method further comprises determining a characterisation profile or model determined from a subset of the classification parameters.

24. A method according to any preceding claim further comprising the step of forwarding a set of characterisation requirements or a requirements request to a third party.

25. A method according to claim 24 further comprising the step of receiving adjustment inputs from the third party.

26. A method according to claim 25, wherein the adjustment inputs from the third party are distinguished by at least one of: a suitable label, a representation of the third party adjustment input with a symbol, icon, identification number, a display of hover-over narrative.

27. A method according to claim 25 or 26, further comprising the step of generating a revised set of characterisation requirements or a requirements request in dependence on the received adjustment input.

28. A method according to any preceding claim, further comprising the step of determining a suitable assessment method or technique in dependence on the set of characterisation requirements.

29. A method according to claim 28 further comprising the step of selecting a plurality of assessment methods or techniques in dependence on the set of requirements.

30. A method according to claim 29 wherein the plurality of assessment methods or techniques are selected in dependence on a plurality of characterisation parameters.

31. A method according to claim 30 wherein the characterisation parameters are competencies.

32. A method according to claim 30 or 31 further comprising associating a characterisation parameter weighting, or an importance class, with each characterisation parameter.

33. A method according to claim 32 wherein the plurality of assessment methods or techniques is further selected according to the characterisation parameter weighting or importance class.

34. A method according to claim 32 or 33 wherein only assessment methods or techniques related to higher weighted or higher classified characterisation parameters being selected.

35. A method according to any of claims 28 to 34 wherein the plurality of selected assessment methods or techniques are limited to a maximum number of assessment methods or techniques.

36. A method according to any of claims 28 to 35 wherein the selected assessment methods or techniques are optimised to relate to a maximum number of characterisation parameters.

37. A method according to any of claims 28 to 36 wherein the selected assessment methods or techniques are optimised to relate to a maximum number of characterisation parameters in a particular importance class.

38. A method according to claim 37 wherein the importance class is 'high' importance.

39. A method according to any of claims 28 to 38 wherein the selected assessment methods or techniques are optimised to relate to a maximum number of combined and weighted characterisation parameters.

40. A method according to claim 39 wherein the selected assessment methods or techniques are optimised by using characterisation parameter weightings.

41. A method according to claim 40 respective assessment method weighting is associated with each assessment method and characterisation parameter.

42. A method according to any preceding claim further comprising the step, prior to receiving the classification parameters, of evaluating a plurality of assessment results to generate a plurality of characterisation parameter results.

43. A method according to claim 42 wherein the assessment results are individual assessment results resulting to a plurality of characterisation parameters.

44. A method according to claim 42 or 43 wherein a plurality of assessment results are weighted and combined to generate a characterisation parameter result.

45. A method according to any preceding claim comprising the step of evaluating a plurality of characterisation parameter results in dependence on a weighting input, said weighting input preferably being said characterisation parameter weighting or importance class.

46. A method according to claim 45 wherein comprising the step of calculating an assessment score in dependence on the plurality of assessment results and weighting inputs.

47. A method according to any preceding claim further comprising at least one of the following functions: administration of candidate entities; submission of invitations to candidates to undertake an assessment; administration of assessment results; collecting, scoring, evaluating and/or reporting of assessment results; presentation of assessment results; and/or other functions relating to a characterisation profile and potential candidate entities.

48. A method of determining a job specification, the method comprising the steps of:

receiving a job classification;

selecting from a database in dependence on the classification a set of matching job profiles;

extracting for each matching job profile a set of assessed competencies;

determining the relative relevance of each of the competencies to the job classification; and

generating a job specification based on the determined relevant competencies.

49. A method according to claim 48 wherein the job classification comprises a standard classification.

50. A method according to claim 49 wherein the classification is the Standard Occupational Classification SOC code as used by 0*NET.

51. A method according to claim 50 wherein the classification is based on a hybrid of standard classifications, wherein the granularity is provided by a first classification, in combination with the set of aggregations of levels provided by a second classification.

52. A method according to any of claims 48 to 51 wherein the competencies are defined according to a competency model of a sufficient number of parameters to allow for compatibility by mapping to an externally supplied competency model.

A method according to claim 52 wherein the competency model is the Universal competency Framework UCF.

54. A method according to any of claims 48 to 53 wherein the relevant competencies for a particular job are be determined from data compiled from assessments of workers employed in the same or similar jobs.

55. A method according to any of claims 48 to 54 wherein the relevance of competencies of a job specification are adjusted according to a set of contextual variables determined by answers provided by the user to questions relating to particulars of the job.

56. A method according to claim 55 wherein the particulars of the job comprise a requirement of the job, the job environment or of the job candidate.

57. A method according to claim 55 or 56 wherein the relevance of competencies of a job specification are adjusted by a moderator to which the job specification is referred.

58. A method according to any preceding claim further comprising the step of determining a raw profile for the entity in direct dependence on the selection of the set of entities in dependence on the classification parameters.

59. A method according to claim 58 wherein the requirements characterisation profile is available alongside the raw profile.

60. A method according to claim 58 or 59 wherein the constructed requirements characterisation profile can be overridden.

61. A method according to claim 58, 59 or 60 wherein the requirements characterisation profile is provided in adjustable slider format and preferably raw profile is available in fixed slider.

62. A method according to claim 61 wherein both an adjustable slider and a fixed slider are provided in conjunction with each other.

63. A method of facilitating the recruitment of a job candidate, the method comprising the steps of:

receiving job classification information;

determining, in dependence on the job classification an initial competency profile.

64. A method of claim 63 further comprising the step of, receiving job context information along with the job classification information.

65. A method according to claim 63 or 64 further comprising the step of refining the competency profile in dependence on the job context information

66. A method according to any of claims 63 to 65 further comprising the step of receiving and incorporating feedback on the competency profile.

67. A method according to any of claims 63 to 66 further comprising the step of determining a suitable assessment for assessing potential candidate for competencies in accordance with the competency profile.

68. An apparatus for determining a requirements characterisation profile for an entity, the apparatus comprising:

means for receiving classification parameters defining a requirement for an entity;

means for selecting, in dependence on the classification parameters, a set of entities from a database of previously assessed entities;

means for retrieving from the database characterisation parameters of the selected set of entities; and

means for constructing, in dependence on the characterisation parameters, a requirements characterisation profile for the entity.

69. An apparatus according to claim 68 further comprising means for carrying out a method according to any of claims 2 to 47.

70. An apparatus for determining a job specification, the apparatus comprising:

means for receiving a job classification;

means for selecting from a database in dependence on the classification a set of matching job profiles;

means for extracting for each matching job profile a set of assessed competencies; means for determining the relative relevance of each of the competencies to the job classification; and

means for generating a job specification based on the determined relevant competencies.

71. An apparatus according to claim 70 further comprising means for carrying out a method according to any of claims 49 to 67.

72. An apparatus according to any claim of 68 to 71 comprising at least one server, for providing access for a client or user either directly or via a computer, to the results of processing.

73. An apparatus according to claim 72 wherein the server provides access via a web interface, to the results of processing.

74. An apparatus according to claim 72 or 73 wherein one or more server functions are combined in a single server.

75. A method substantially as described herein with reference to the accompanying figures.

76. An apparatus substantially as described herein with reference to the accompanying figures.