Event Information
Blended ContentOther presentations in this group:
The research includes the perspective to analyze the educators’ data and the framework to train AI model.
Firstly, this research utilizes a tree metaphor perspective to analyze the educators’ data to construct them into major three components: a) roots: fundamental part of educator’s identity b) trunk: teaching-related information and knowledge, and c) leaf or fruit: educators’ teaching outcome.
Secondly, as for the framework to train AI model, we use Recurrent Neural Network (RNN) model (Sherstinsky, 2020) as the framework. A recurrent neural network (RNN) is a type of deep learning model that processes sequential data to produce a sequential output. Using RNN, we train machine learning algorithms to predict the major characteristics of each teacher, thus they can have AI technological support in having on-going understanding of identity development.
For training AI, the research design follows the interview with educators and then uses the interview data to create and train an AI model. The goal of the AI model is to identify patterns and insights from the collected data, specifically teacher responses, to categorize teacher characteristics. Teacher responses were gathered by interview. At the interviews, ten educators in Texas have interview questions to understand their identity as teachers. For example, a) what part of identity is important for you as an educator? b) What beliefs, philosophies, and values are important to you? and c) What is foundational and central for you as educators, which is not tangible and visible? d) What information, knowledge, and pedagogical information do you have?
The teachers’ responses, gathered from the questions above, undergo a meticulous preprocessing phase. This involves the removal of stop words, symbols, and numbers from the responses, as well as the conversion of all words to lowercase. We then apply Word2Vec, a powerful tool that generates word embeddings, transforming words into continuous vector representations that capture their semantic meanings. The data is organized to meet the requirements of our chosen unsupervised learning algorithm. We opted for clustering algorithms to group the data into specified clusters based on similarity, allowing us to analyze similarities in teacher responses and characterize them effectively. We also employ a Recurrent Neural Network (RNN) model, specifically a Long Short-Term Memory (LSTM) variant, which processes words sequentially while maintaining an internal memory of previously encountered inputs. LSTMs enhance traditional RNNs by incorporating a mechanism that allows them to access information from any previous timestep. This capability creates a long-term memory structure that retains all past inputs along with their timestamps. The LSTM introduced greater complexity to our network, enabling it to uncover more meaningful relationships between inputs and their contextual timing. The training model is then evaluated and iterated to fit our objective and application.
The result of this research is an AI model we train to predict the major characteristics of teacher identity. This AI can be used to provide ongoing professional development support for educators to share their thoughts about who they are as educators. The AI will aid them in characterizing their sharing into major components related to teacher identity. This AI can be incorporated into teacher education programs. Understanding the current condition of teacher identity can provide insight for educators to set the areas within teacher identity to develop further to become the educators they envision.
For the professional development of educators, ongoing reflection on their practice is essential to improve teaching. This research outcome about the AI model is a collaborative tool that can be used to support educators in reflecting on and understanding their current identity as educators. It enables them to understand their unique teacher identity and work together to identify growth areas and improve their teaching practice for the benefit of their students.
As for the academic contribution, this research would extend existing literature about how researchers can utilize AI technology to support teachers in becoming the educators they envision and improving their teaching practice. This research aligns with the 2024 ISTE conference, which will present diverse ideas on using technology to explore opportunities to make learning more meaningful. Specifically, this research will provide the conference participants with information about AI technology to explore new methods of professional development opportunities.
• Chu, Y. (2021). Preservice teachers learning to teach and developing teacher identity in a teacher residency. Teaching Education, 32(3), 269-285.
• Eğmir, E., & Çelik, S. (2019). The educational beliefs of pre-service teachers as an important predictor of teacher identity. International Journal of Contemporary Educational Research, 6(2), 438-451.
• Jenlink, P. M. (Ed.). (2021). Understanding teacher identity: The complexities of forming an identity as professional teacher. Rowman & Littlefield.
• Harlow, A., & Cobb, D. J. (2014). Planting the seed of teacher identity: Nurturing early growth through a collaborative learning community. Australian Journal of Teacher Education (Online), 39(7), 70-88.
• Sherstinsky, A. (2020). Fundamentals of recurrent neural network (RNN) and long short-term memory (LSTM) network. Physica D: Nonlinear Phenomena, 404, 132306.
• Ruohotie-Lyhty, M., & Moate, J. (2016). Who and how? Preservice teachers as active agents developing professional identities. Teaching and teacher education, 55, 318-327.
Back
Trips and Tours