Page 95 - Ai_C10_Flipbook
P. 95

Following a well-defined AI project cycle ensures that every stage of the AI project is methodical and minimises
                 unforeseen issues. The following figure shows six stages of the AI project cycle:
                 The description of these six stages is as follows:

                  Stage 1   Problem Scoping

                 The first stage of an AI project cycle is Problem Scoping, where the problem is clearly identified, and a vision for
                 addressing the same is developed. It is a crucial step where the focus is thoroughly on understanding the problem
                 by considering the various factors that influence it, and determining how AI technology can provide a solution.
                 To achieve a comprehensive understanding, this phase emphasises the use of the 4W’s: Who, What, Where, and
                 Why. This approach ensures that all critical components of the problem are clearly defined, aligning stakeholders
                 and team members.
                    • Who: This step identifies who will be affected from the AI solution, as well as any stakeholders involved in the
                   project. It considers the target audience, users, and decision-makers.

                    • What: This step defines the specific problem or challenge that needs to be addressed with AI. It outlines the
                   goals and the desired outcome of the project.
                    • Where: This step focuses on where the AI solution will operate or be implemented. It could refer to the technical
                   environment, the geographical location, or the specific domain.

                    • Why: This step explores the reason behind solving the problem. It looks at the value and impact that solving
                   the problem will have for the business, users, or society.
                 These questions serve as a structured approach to ensure that all relevant aspects of the problem are considered,
                 thereby facilitating a more effective and efficient analysis and solution development.

                 Once  you’ve  completed the 4W's                      4W Problem Statement Template
                 problem  canvas,  the  next  step  is  to   Our                  [stakeholders]                 Who
                 consolidate all the information into a
                                                        has a problem that        [issue, problem, need]         What
                 single,  comprehensive  template.  The
                                                        when/while                [context, situation]           Where
                 problem statement template allows
                                                        An ideal solution would   [benefit of solution for them]  Why
                 you to compile all the key details into
                 one format, providing a clear reference point for future use. Problem Statement Template is a structured format
                 to articulate the problem clearly. It ensures that the problem is well-defined and understood by all stakeholders.
                 Below is a template with designated spaces to fill in details aligned with your objectives.

                  Stage 2   Data Acquisition

                 The next stage in the AI project cycle is known as data acquisition. This stage involves gathering raw data, which
                 is essential for referencing or performing analysis that will guide the project. The process of data acquisition
                 encompasses the collection of a wide range of data types, including text, numerical values, images, videos, and
                 audio. These various forms of data can be sourced from multiple places such as the internet, academic journals,
                 newspapers, and other relevant publications or databases. The goal of data acquisition is to capture accurate and
                 valuable information that reflects real-world scenarios. This collected data serves as a foundation of the project
                 that provides valuable insights and enabling the improvement of the project’s performance and the development
                 of more precise AI solutions.

                  Stage 3   Data Exploration
                 Data exploration is a crucial step that involves analysing large volumes of data to uncover meaningful patterns,
                 trends, and relationships using various data visualisation and statistical techniques. By transforming raw data into

                                                            Revisiting AI Project Cycle & Ethical Frameworks for AI   93
   90   91   92   93   94   95   96   97   98   99   100