Completeness in dependency networks

1. A computer program stored on a non-transitory computer-readable medium for automated customized document generation where inclusion of document content of a template is based on evaluation of rules, the rules being evaluated using values assigned to content variables of the template, wherein the values are based on responses to questions of a questionnaire, comprising:

logic which, for each question of the questionnaire for which a response is received, determines a value based on the response and assigns the value to a corresponding content variable of the template associated with the question;

logic which, after the questionnaire has been completed, and where received responses are insufficient to evaluate all rules, determines:

which content variables of the template are relevant; and

whether all relevant content variables of the template are in the dataset of content variables for which values have been assigned, including:

that the variables form a well-formed dependency network wherein the dependency network is said to be well-formed if, and only if, the following condition is satisfied:

for every arc which involves two or more source nodes, either:

(i) none of the source nodes are themselves target nodes of arcs, or

(ii) all of the source nodes are target nodes of other arcs, and all those arcs are tagged with same condition;

wherein each node represents a variable and each arc represents a connection between source nodes and target nodes; and

given the well-formed dependency network and the dataset, the variable is defined as relevant if, and only if:

(i) the variable is not the target node of any arcs, or

(ii) it is the target node of an arc, and all the source node variables are relevant, and the conditions do evaluate to true or indefinite; and

wherein a dataset is defined as complete if, and only if, all relevant variables are in the dataset and indicating where a dataset is not complete by listing those variables which are missing from the dataset; and

logic which generates a fully customized document from the template if it is determined that all relevant content variables of the template are in the dataset.

2. The computer program of claim 1 including using the values of the variables to determine which items of information are incorporated into a customised document.

3. The computer program of claim 2 including generating the customised document which is a legal document.

4. A method for producing a customised document for which inclusion of document content of a template is based on evaluation of rules, the rules being evaluated using values assigned to content variables of the template, wherein the values are based on responses to questions of a questionnaire, comprising the steps of:

completing a questionnaire to define a dataset of content variables for which values have been assigned;

determining which content variables of the template are relevant;

determining whether all relevant variables of the template are in the dataset, including:

determining that the variables form a well-formed dependency network wherein the dependency network is said to be well-formed if, and only if, the following condition is satisfied:

for every arc which involves two or more source nodes, either:

(i) none of the source nodes are themselves target nodes of arcs, or

(ii) all of the source nodes are target nodes of other arcs, and all those arcs are tagged with same condition;

wherein each node represents a variable and each arc represents a connection between source nodes and target nodes; and

given the well-formed dependency network and the dataset, the variable is defined as relevant if, and only if:

(i) the variable is not the target node of any arcs, or

(ii) it is the target node of an arc, and all the source node variables are relevant, and the conditions do evaluate to true or indefinite; and

wherein a dataset is defined as complete if, and only if, all relevant variables are in the dataset and indicating where a dataset is not complete by listing those variables which are missing from the dataset; and

where received responses are insufficient to evaluate all rules, generating a fully customised document if all relevant variables of the template are in the dataset.

5. The method of claim 4 including using the values of the variables to determine which items of information are incorporated into a customized document.

6. The method of claim 5 including generating the customized document which is a legal document.