Intelligent and Interactive Chatbot Based on the Recommendation Mechanism to Reach Personalized Learning

Intelligent and Interactive Chatbot Based on the Recommendation Mechanism to Reach Personalized Learning

Ching-Bang Yao, Yu-Ling Wu
DOI: 10.4018/IJICTE.315596
Article PDF Download
Open access articles are freely available for download

Abstract

With the impacts of Covid-19 epidemic, e-learning has become a popular research issue. Therefore, how to upgrade the interactivity of e-learning, and allow learners to quickly access personalized and popular learning information from huge digital materials, is very important. However, chatbots are mostly used in automation, as well as simple occasions of general standard question and answer. But to solve the different problems of e-learners in the learning process, chatbots are used to filter the blind spots of learners and to provide further relevant information, so that e-learning can improve in efficiency and interactivity. This study utilizes AI, two-stage Bayesian algorithm, and crawler technology to provide customized learning materials according to learner's current learning situation. The experimental results show that this research system can indeed correctly understand and judge the blind spots of digital learners, and effectively find the relevant e-learning and video information. The accuracy rate reaches nearly 90%.
Article Preview
Top

Introduction

The Cross-domain learning has become an irresistible trend in recent years. The e-learning itself means that the learning materials and resources in other professional fields can be viewed at any time, and it is not limited to any time or place. The e-learning has developed into an important learning method and channel for cross-domain learning in other professional fields. However, how to create an excellent digital learning platform to allow e-learners to reduce their difficulties and barriers in different subjects or college majors will be the most important key factor affecting the effect of e-learning.

This research utilized courses in learning to cook as an example, with the concept of healthy eating gaining popularity during the Coronavirus disease (COVID-19) lockdowns. However, despite the various recipes and teaching videos on the Internet, it can be impossible to solve the various problems or exceptions encountered by the cook owing to one or more technical gaps, leading to the failure of the finished product. The proper use of an e-learning platform and artificial intelligence technologies (Ayodele et al., 2011; Wahyono, 2019) could immediately provide solutions to actual problems for users learning to cook. This could also provide more detailed teaching explanations and instructions when encountering such problems, thereby effectively addressing the key blind spots in cooking. That is, when learners watch online teaching videos (Carver et al., 1999; Salehi et al., 2013), they can ask for more detailed answers in time by entering keywords to the robot teaching assistant when they encounter problems (Chen et al., 2008; Chen, 2011). The finished dish will not only be more fulfilling but will also make the entire e-learning process more interesting. This research also used two-stage Bayesian algorithms and chatbots to interact with learners and created an intelligent e-learning platform to help them effectively and instantly solve the difficulties encountered in the learning process (Ramaswami & Rathinasab, 2012; Yao, 2017; Ogata & Yano, 2004). Simultaneously, this approach could bring the real-time puzzle function of a virtual teacher to e-learning (e.g., to solve the above problems) (Chen & Li, 2010), attract increased user attention to active learning, and enhance the effects of e-learning (Chu et al., 2004; Cormac & Siobhan, 2008).

Many people are interested in learning about various topics after they graduate or outside of their original majors, but they often do not have sufficient time to take these professional courses, owing to tremendous factors, such as work, school, and family. Therefore, the e-learning has become the gospel for students in online learning (Obasa et al., 2013). Still, the current popular research topics in e-learning mainly focus on the production and provision of digital teaching materials and how to automatically provide additional digital teaching materials to learners. As a result, learners are often faced with a huge amount of digital learning materials, but they do not know where they should start learning at present or where they really want to learn. Therefore, e-learning often lacks the proper digital textbooks to give students when they encounter problems and is unable to provide direct guidance (like a real person could) or find the crux of the problem. Moreover, e-learning cannot quickly understand the student’s situation based on the student’s current encountered problems. Because these problems in the learning process (learning blind spots) cannot be solved immediately, the learning process will either be stuck and cannot continue, or the students even give up learning.

Accordingly, e-learning often lacks interactivity and an understanding of the learners’ actual learning and absorption status during the learning process (Chen, 2011). It is necessary, therefore, to rely on learners to select courses that meet their own learning progress from a large number of digital textbooks. But this requirement is difficult for most digital learners. The e-learners really need to get instant solutions and help in the process of distance learning.

Complete Article List

Search this Journal:
Reset
Volume 20: 1 Issue (2024)
Volume 19: 1 Issue (2023)
Volume 18: 3 Issues (2022)
Volume 17: 4 Issues (2021)
Volume 16: 4 Issues (2020)
Volume 15: 4 Issues (2019)
Volume 14: 4 Issues (2018)
Volume 13: 4 Issues (2017)
Volume 12: 4 Issues (2016)
Volume 11: 4 Issues (2015)
Volume 10: 4 Issues (2014)
Volume 9: 4 Issues (2013)
Volume 8: 4 Issues (2012)
Volume 7: 4 Issues (2011)
Volume 6: 4 Issues (2010)
Volume 5: 4 Issues (2009)
Volume 4: 4 Issues (2008)
Volume 3: 4 Issues (2007)
Volume 2: 4 Issues (2006)
Volume 1: 4 Issues (2005)
View Complete Journal Contents Listing