Handbook_of_Artificial_Intelligence_in_Education_----_(2._The_history_of_artificial_intelligence_in_education_-_the_first_qua...)
Introduction to AIED
Definition: Artificial intelligence in education (AIED) is an interdisciplinary field.
Roots in: Computer science, cognitive science, education, and social sciences.
Goal: Create software that enhances learning experiences for individuals.
Historical Attraction: Early researchers were drawn to the education domain for several reasons:
Interesting Microworld: Education provides conditions that can yield better computational models than robotics or expert systems.
Cognitive Modelling: Cognitive scientists valued education as a platform to demonstrate cognitive theories.
Applied Learning Technology: Education researchers aimed to innovate learning environments through technology.
Evolution as a Field: Over time, diverse perspectives from these disciplines converged into a coherent applied field focusing on advanced computational techniques in educational contexts.
Historical Context of AIED
Prehistoric Contributions: Before the digital age, several pioneers contributed to educational machinery:
Pressey (1926): Developed special-purpose machines for multiple-choice questions.
Skinner (1954): Influenced by Pressey, he conceptualized teaching machines that graded answers.
Behaviorist Foundations: Early automated teaching approaches were framed within behaviorist psychology.
Emergence of Computer-Assisted Instruction (CAI)
In the 1960s, CAI systems were born, consisting of databases of educational material with feedback mechanisms.
Branching: Systems could adapt content based on student performance, akin to AIED's adaptive structure.
Limitations: The need to pre-define every possible student response made CAI systems complex.
Generative CAI
Late 1970s introduced generative CAI, capable of creating its own questions based on the student’s progress.
SCHOLAR System: Recognized as one of the first AIED systems, facilitating inquiry-based learning through Socratic dialogue focused on South American geography.
Knowledge Representation: Incorporated a semantic network for better fact retrieval and inference making.
The First Era of AIED Research (1970 - 1982)
Intelligent CAI Systems (ICAI) emerged during the golden years of CAI and AI:
Systems were expected to reason and understand knowledge deeply, leading to more effective learning aids.
Major Contributions and Systems
SCHOLAR (1970): The pioneering ICAI system that utilized AI principles in education.
The WHY System: Captured Socratic tutoring methodologies for educational applications.
SOPHIE: Focused on hands-on electronic circuit troubleshooting.
Problem Generation: Research by Ramani and Newell explored problem generation strategies using semantic networks.
Pedagogical Approaches
Systems aimed to emulate a “guide on the side” rather than a traditional teacher, focusing on adaptability and responsiveness to learners.
The Second Era of AIED Research (1982 - 1995)
The Sleeman and Brown (1982) edited collection was crucial in establishing AIED as a distinct discipline.
Intelligent Tutoring System Architecture: A framework for structuring sophisticated educational software.
Modules involved: Domain knowledge, student modeling, pedagogical strategies, and communication components.
Cognitive Tutoring Paradigm
Developed extensively under John Anderson at Carnegie Mellon University:
Skill Acquisition: Systems aimed to teach students essential problem-solving skills using cognitive principles and production rules.
Immediate Feedback: Based on model tracing, enabling timely error correction, though sometimes rigid.
Expanding Parameters and Capabilities
AIED began integrating diverse pedagogical strategies and exploring constructivist philosophies, shifting towards greater learner autonomy and exploration.
Logistics and Research Integration: AIED reflected on the need for long-term commitment to certain paradigms for richer academic exploration.
Evaluation and Real-World Deployment
Evaluating AIED systems proved challenging due to the dynamic nature of these educational tools in fluid learning environments.
Commercialization Effects: Successful real-world deployments began to demonstrate both promise and hurdles in operationalizing these AIED technologies.
Conclusion: Lessons from 25 Years of AIED Research
Historical Awareness: Understanding foundational developments can yield insights for current research.
Diversity in Approaches: Welcoming varied methodologies enriches the discourse in AIED.
Adaptive Learning: Continual evolution in learner modeling and student interactions promises to redefine educational landscapes.