Method and apparatus for selecting taxpayer audits

An audit selection method comprising:

• creating an underreporting model indicating probability of an assessment being made against respective taxpayers from historical data;
• creating an analytical space for analyzing the taxpayers, the analytical space being based on the underreporting model;
• positioning taxpayers within the analytical space;
• analyzing factors within the analytical space based on historical data;
• using the analyzed factors to predict audit return on auditors' time when auditing taxpayers;
• forecasting assessments against taxpayers based on the predicted audit return;
• and
• formulating a basis for audit selection from the forecasted assessments, wherein
• the underreporting model scores taxpayers based on historical data to provide respective taxpayers with a score indicating probability of an assessment being made against the taxpayer if an audit of the taxpayer is conducted, and assigns taxpayers into categories so that taxpayers with the same score can have different probabilities of an assessment based on the category of the taxpayer,
• the analytical space is a graph having a first axis indicating score and a second axis indicating category, the graph having cells at each intersection of score and category,
• said analyzing factors comprises
• recalculating the probability of making an assessment within each cell, based on the actual taxpayers in that cell,
• performing a regression for each category based on the recalculated probabilities, to determine an equation for each category which forecasts probabilities that an assessment will be made for taxpayers in that category,
• calculating an average assessment amount for cases in each cell, and
• determining the time taken to perform an audit for each cell, and
• said using the analyzed factors uses the equations, calculated average assessment amounts and determined time taken to perform an audit, to predict audit return.

An audit selection method comprising:

• scoring taxpayers based on historical data to provide respective taxpayers with a score indicating a probability of an assessment being made against the taxpayer if an audit of the taxpayer is conducted;
• assigning taxpayers into categories so that taxpayers with the same score can have different probabilities of an assessment based on the category of the taxpayer;
• creating an analytical space for analyzing the taxpayers, the analytical space defined by the taxpayer score and category;
• positioning taxpayers within the analytical space based on their score and category;
• analyzing factors within the analytical space based on historical data;
• using the analyzed factors to predict audit return on auditors' time when auditing taxpayers;
• forecasting assessments against taxpayers based on the predicted audit return on auditors' time;
• and
• formulating a basis for audit selection from the forecasted assessments, wherein
• the analytical space is a graph having a first axis indicating score and a second axis indicating category, the graph having cells at each intersection of score and category,
• said analyzing factors comprises
• recalculating the probability of making an assessment within each cell, based on the actual taxpayers in that cell,
• performing a regression for each category based on the recalculated probabilities, to determine an equation for each category which forecasts probabilities that an assessment will be made for taxpayers in that category,
• calculating an average assessment amount for cases in each cell, and
• determining the time taken to perform an audit for each cell, and
• said using the analyzed factors uses the equations, calculated average assessment amounts and determined time taken to perform an audit, to predict audit return on auditors' time.

An audit selection method comprising:

• scoring taxpayers based on historical data to provide respective taxpayers with a score indicating a probability of an assessment being made against the taxpayer if an audit of the taxpayer is conducted;
• assigning taxpayers into categories so that taxpayers with the same score can have different probabilities of an assessment based on the category of the taxpayer;
• creating an analytical space, the analytical space being a graph having a first axis indicating score and a second axis indicating category, the graph having cells at each intersection of score and category;
• positioning taxpayers within the analytical space based on their score and category;
• recalculating the probability of making an assessment within each cell of the graph, based on the actual taxpayers in that cell;
• performing a regression for each category based on the recalculated probabilities, to determine an equation for each category which forecasts probabilities that an assessment will be made for taxpayers in that category;
• calculating an average assessment amount for cases in each cell of the graph;
• determining the time taken to perform an audit for each cell of the graph;
• using the equations, calculated average assessment amounts and determined time taken to perform an audit, to predict audit return on auditors' time;
• forecasting assessments against taxpayers based on the predicted audit return on auditors' time;
• and
• formulating a basis for audit selection from the forecasted assessments.

An audit selection method comprising:

• selecting taxpayers to audit to maximize a total audit assessment amount, said selecting taxpayers comprising
• scoring taxpayers based on historical data to provide respective taxpayers with a score indicating probability of an assessment being made against the taxpayer if an audit of the taxpayer is conducted;
• assigning taxpayers into categories so that taxpayers with the same score can have different probabilities of an assessment based on the category of the taxpayer;
• creating an analytical space for analyzing the taxpayers, the analytical space being a graph having a first axis indicating score and a second axis indicating category, the graph having cells at each intersection of score and category;
• positioning taxpayers within the analytical space based on their score and category;
• analyzing factors within each cell of the analytical space based on actual taxpayers within that cell and historical data via the use of regression;
• using the analyzed factors to predict audit return on auditors' time when auditing taxpayers;
• forecasting assessments against taxpayers based on the predicted audit return on auditors' time;
• and
• formulating a basis for audit selection from the forecasted assessments, wherein said analyzing factors produces an average assessment amount for taxpayers in each cell and time taken to perform an audit for taxpayers in each cell.

An audit selection method comprising:

• selecting taxpayers to audit to maximize a total audit assessment amount, said selecting taxpayers comprising
• scoring taxpayers based on historical data to provide respective taxpayers with a score indicating probability of an assessment being made against the taxpayer if an audit of the taxpayer is conducted;
• assigning taxpayers into categories so that taxpayers with the same score can have different probabilities of an assessment based on the category of the taxpayer;
• creating an analytical space for analyzing the taxpayers, the analytical space being a graph having a first axis indicating score and a second axis indicating category, the graph having cells at each intersection of score and category;
• positioning taxpayers within the analytical space based on their score and category;
• analyzing factors within each cell of the analytical space based on actual taxpayers within that cell and historical data via the use of regression;
• using the analyzed factors to predict audit return on auditors' time when auditing taxpayers;
• forecasting assessments against taxpayers based on the predicted audit return on auditors' time;
• and
• formulating a basis for audit selection from the forecasted assessments, wherein said analyzing factors comprises:
• recalculating the probability of making an assessment within each cell, based on the actual taxpayers in that cell, and
• performing an analysis for each category based on the recalculated probabilities, to determine an equation for each category which forecasts probabilities that an assessment will be made for taxpayers in that category.

An audit selection method comprising:

• selecting taxpayers to audit to maximize a total audit assessment amount, said selecting taxpayers comprising
• scoring taxpayers based on historical data to provide respective taxpayers with a score indicating probability of an assessment being made against the taxpayer if an audit of the taxpayer is conducted;
• assigning taxpayers into categories so that taxpayers with the same score can have different probabilities of an assessment based on the category of the taxpayer;
• creating an analytical space for analyzing the taxpayers, the analytical space being a graph having a first axis indicating score and a second axis indicating category, the graph having cells at each intersection of score and category;
• positioning taxpayers within the analytical space based on their score and category;
• analyzing factors within each cell of the analytical space based on actual taxpayers within that cell and historical data via the use of regression;
• using the analyzed factors to predict audit return on auditors' time when auditing taxpayers;
• forecasting assessments against taxpayers based on the predicted audit return on auditors' time;
• and
• formulating a basis for audit selection from the forecasted assessments, wherein said analyzing factors comprises recalculating probability of making an assessment within each cell, based on actual taxpayers in that cell, and said using the analyzed factors predicts audit return in accordance with the recalculated probabilities.

An audit selection method comprising:

• selecting taxpayers to audit to maximize a total audit assessment amount, said selecting taxpayers comprising
• scoring taxpayers based on historical data to provide respective taxpayers with a score indicating probability of an assessment being made against the taxpayer if an audit of the taxpayer is conducted;
• assigning taxpayers into categories so that taxpayers with the same score can have different probabilities of an assessment based on the category of the taxpayer;
• creating an analytical space for analyzing the taxpayers, the analytical space being a graph having a first axis indicating score and a second axis indicating category, the graph having cells at each intersection of score and category;
• positioning taxpayers within the analytical space based on their score and category;
• analyzing factors within each cell of the analytical space based on actual taxpayers within that cell and historical data via the use of regression;
• using the analyzed factors to predict audit return on auditors' time when auditing taxpayers;
• forecasting assessments against taxpayers based on the predicted audit return on auditors' time;
• and
• formulating a basis for audit selection from the forecasted assessments, wherein said analyzing factors comprises performing an analysis to determine an equation for each category which forecasts that an assessment will be made for taxpayers in that category, and said using the analyzed factors predicts audit return in accordance with the equations.

An audit selection comprising:

• selecting taxpayers to audit to maximize a total audit assessment amount, said selecting, taxpayers comprising
• scoring taxpayers based on historical data to provide respective taxpayers with a score indicating probability of an assessment being made against the taxpayer if an audit of the taxpayer is conducted,
• assigning taxpayers into categories so that taxpayers with the same score can have different probabilities of an assessment based on the category of the taxpayer,
• creating an analytical space for analyzing the taxpayers, the analytical space defined by the taxpayer score and category,
• positioning taxpayers within the analytical space based on their score and category,
• analyzing factors within the analytical space based on historical data,
• using the analyzed factors to predict audit return on auditors' time when auditing taxpayers,
• forecasting assessments against taxpayers based on the predicted audit return on auditors' time, and
• formulating a basis for audit selection from the forecasted assessments, wherein
• the analytical space is a graph having a first axis indicating score and a second axis indicating category, the graph having cells at each intersection of score and category,
• said analyzing factors comprises
• recalculating, from historical data, the probability of making an assessment within each cell, based on the actual taxpayers in that cell,
• performing a regression analysis for each category based on the recalculated probabilities, to determine an equation for each category which forecasts probabilities that an assessment will be made for taxpayers in that category,
• calculating an average assessment amount for cases in each cell, and
• determining the time taken to perform an audit for each cell, and
• said using the analyzed factors uses the equations, calculated average assessment amounts and determined time taken to perform an audit, to predict audit return.

An audit selection method comprising:

• creating an underreporting model indicating probability of an assessment being made against respective taxpayers from historical data;
• creating an analytical space for analyzing the taxpayers, the analytical space being based on the underreporting model;
• positioning taxpayers within the analytical space;
• analyzing factors within the analytical space based on historical data;
• using the analyzed factors to predict audit return on auditors' time when auditing taxpayers;
• forecasting assessments against taxpayers based on the predicted audit return;
• and
• formulating a basis for audit selection from the forecasted assessments, wherein
• the underreporting model scores taxpayers based on historical data to provide respective taxpayers with a score indicating probability of an assessment being made against the taxpayer if an audit of the taxpayer is conducted, and assigns taxpayers into categories so that taxpayers with the same score can have different probabilities of an assessment based on the category of the taxpayer,
• the analytical space is a graph having a first axis indicating score and a second axis indicating category, the graph having cells at each intersection of score and category,
• said analyzing factors comprises
• recalculating the probability of making an assessment within each cell, based on the actual taxpayers in that cell,
• performing a regression analysis for each category based on the recalculated probabilities, to determine an equation for each category which forecasts probabilities that an assessment will be made for taxpayers in that category,
• calculating an average assessment amount for cases in each cell, and
• determining the time taken to perform an audit for each cell, and
• said using the analyzed factors uses the equations, calculated average assessment amounts and determined time taken to perform an audit, to predict audit return.

An audit selection method comprising:

• scoring taxpayers based on historical data to provide respective taxpayers with a score indicating a probability of an assessment being made against the taxpayer if an audit of the taxpayer is conducted;
• assigning taxpayers into categories so that taxpayers with the same score can have different probabilities of an assessment based on the category of the taxpayer;
• creating an analytical space for analyzing the taxpayers, the analytical space defined by the taxpayer score and category;
• positioning taxpayers within the analytical space based on their score and category;
• analyzing factors within the analytical space based on historical data;
• using the analyzed factors to predict audit return on auditors' time when auditing taxpayers;
• forecasting assessments against taxpayers based on the predicted audit return on auditors' time;
• and
• formulating a basis for audit selection from the forecasted assessments, wherein
• the analytical space is a graph having a first axis indicating score and a second axis indicating category, the graph having cells at each intersection of score and category,
• said analyzing factors comprises
• recalculating the probability of making an assessment within each cell, based on the actual taxpayers in that cell,
• performing a regression analysis for each category based on the recalculated probabilities, to determine an equation for each category which forecasts probabilities that an assessment will be made for taxpayers in that category,
• calculating an average assessment amount for cases in each cell, and
• determining the time taken to perform an audit for each cell, and
• said using the analyzed factors uses the equations, calculated average assessment amounts and determined time taken to perform an audit, to predict audit return on auditors' time.