• Australian Taxation Office


    With a rapidly increasing number of decision-making processes to be completed within short time frames, the Australian Taxation Office (ATO) needs effective software to assist in developing expert systems to expedite these processes.

    We have been using XpertRule to develop an expert system to assist in applying policy to requests from tax practitioners for additional time to lodge their client’s income tax returns.

    Expert systems technology was introduced to bring consistency to the decision making process, as well as speed it up through semi-automation. We wanted to ensure that, with a given set of circumstances, each practitioner would be given the same decision about their application. Links are being tested between the expert system and the ATO’s mainframe so that complete automation may be achieved for the majority of applications.

    We plan to have systems deployed on the intranet and/or Internet sites for easy access to all staff and/or clients. A key strategy of the ATO is to encourage and enable as many clients as possible to use self-help. The use of expert systems is a key component in achieving this.

    Improved objective decision making

    Extensive analysis was done with subject matter experts to develop the business rules before building the expert system. As there is a significant element of subjectivity in this kind of decision-making, the most difficult part of developing the business rules was determining what factors to take into consideration when making a decision.

    The most obvious one was the reason given. A comprehensive list of reasons used was compiled from archived applications and each was given a weighting. Tolerance levels were set to categorise the reasons as “disaster”, “exceptional”, “normal” or “trivial” depending on the total weighting score.

    The decision tree is branched at this point, and only additional relevant information is requested. For example, if the reason is “disaster,” then as long as the requested extension date is before the end of the financial year, the extension will be granted.

    Other factors taken into consideration are:

    • Program – each practitioner is set to one of five lodgment programs. The due date of a return is dependent on the practitioner’s lodgment program.
    • History – the practitioner’s lodgment patterns are assessed as being “good” or “bad”. The taxpayer’s lodgment history is given a score dependent on whether previous returns have been lodged on time.
    • Proposition – this compares the date the return is due to be lodged with the date the practitioner is proposing. It is classified as “acceptable” or “unacceptable”. The length of time varies in accordance with the grounds, history and other related factors.
    • Timing – if the agent has applied too early (that is, we believe they still have sufficient time to prepare and lodge the return by the due date) or have left the request until the last day, the application will be rejected. Exceptions have been built into the system in the case of “disasters”.

    National implementation

    The expert system was tested in one branch office and a post implementation review recommended national implementation. Work is underway to refine and simplify the business rules to achieve better and more equitable outcomes.

    Electronic forms

    Information is currently entered by ATO staff working from written requests. On completion of a case, XpertRule generates a standard letter for the operator to mail to the practitioner.

    Development is underway to capture the information needed by the expert system using an electronic form. The practitioner will fill in the form on their PC and transmit it to the ATO over a secure link. The data is then imported by the expert system to make the decision.

    Operators will only need to enter any data that is stored on other ATO business systems. On completion, the case will be transmitted to a workflow management system for archiving.

     

  • Try Now & Live Demos