اولویت‌بندی ابزارهای هوش مصنوعی چت جی پی تی، جمینای، کوپایلوت و بینگ در حوزه مدیریت دانش

Introduction 
AI chatbots have emerged as effective tools by simulating human conversations and providing accurate and optimized responses, increasingly replacing traditional search methods and gaining popularity among users. On the other hand, the ability of AI chatbots to appropriately respond to user queries in the field of knowledge management is crucial for facilitating access to information, improving decision-making processes, enhancing organizational efficiency, and finding accurate and relevant answers to support learning and knowledge updating within organizations. The aim of this study is to prioritize AI chatbots in the field of knowledge management based on usefulness, information quality, and user-friendliness, from the perspective of experts.
 
Mothodology
This research is an applied study conducted using a mixed-methods approach. Given that the research is carried out in two phases – the first phase involving the selection of research questions to pose to artificial intelligence robots, and the second phase involving the prioritization of responses from AI chatbots to rank the four AI chatbots under investigation based on the examined criteria – two methods, content analysis and the multiple-criteria decision-making (MCDM) technique, have been employed based on the structure of each of these sections. To select the research questions for querying the AI robots, the Google Trends database was utilized. In this database, search terms from global searches conducted between 2004 and June 20, 2024, were collected. Subsequently, considering the necessity of formulating these terms as questions for the AI chatbots, the terms were transformed into interrogative sentences. Finally, after removing several irrelevant terms, five questions – "What is knowledge management?", "Why is knowledge management important?", "What is knowledge?", "What are the processes of knowledge management?", and "Please introduce some tools and techniques of knowledge management" – were systematically searched in the AI chatbots ChatGPT (January 24 version), Gemini, Bing, and Copilot. For this purpose, all browser data was completely cleared before initiating the search, and separate accounts were created for interaction with each AI chatbot to ensure significant differentiation. Each request was processed in a separate chat page to ensure the isolation and optimization of the analytical process. For the prioritization of the AI chatbots' responses, considering the use of multiple criteria in this research, the Analytic Hierarchy Process (AHP) method was employed. Subsequently, using the pairwise comparison method via questionnaires, the questions and saved responses from the AI chatbots were provided to ten experts in the field of knowledge management. Furthermore, three main criteria – "information quality," "usability," and "user-friendliness" – were considered for evaluating the responses based on a review of prior research, allowing the experts to assess the responses according to these three criteria. Additionally, the experts were asked to provide their assessment of the importance of the three criteria. Expert Choice software version 11 was used to design the hierarchical tree.
 
Findings
In comparing the responses of four AI chatbots to questions about knowledge management based on the research indicators, Gemini ranks first with a weight of 0.418, followed by ChatGPT in second place with 0.315, Bing in third with 0.167, and Copilot in fourth with 0.1. According to the research findings, when evaluating the responses of the four AI chatbots based on the information quality criterion, Gemini again ranks first with a weight of 0.525, ChatGPT second with 0.224, Bing third with 0.158, and Copilot fourth with 0.093. This indicates that Gemini provided more appropriate responses to knowledge management questions in terms of information quality. In terms of practicality, ChatGPT ranks first with a weight of 0.379, followed by Gemini in second with 0.357, Bing in third with 0.167, and Copilot in fourth with 0.097. This suggests that, from the experts’ point of view, the information provided by ChatGPT and Gemini is more practical than that of the other chatbots examined. With respect to the user-friendliness criterion, Gemini once again ranks first with a weight of 0.395, followed by ChatGPT with 0.264, Bing with 0.192, and Copilot with 0.149. Therefore, Gemini’s responses are considered more user-friendly compared to the other chatbots in the study. Finally, in the overall prioritization of the research criteria—based on comparing the responses of the four AI chatbots to knowledge management questions—the experts ranked the indicators as follows: practicality ranked first with a weight of 0.505, information quality second with 0.397, and user-friendliness, with a significantly lower weight of 0.097, ranked third. In conclusion, practicality was considered the most important criterion, while user-friendliness was deemed the least important.
Discussion and Conclusion
Previous studies indicate a growing trend in the use of chatbots and suggest that they enhance the user experience through a more conversational interface. However, the quality of chatbot responses largely depends on users' skill in formulating appropriate questions and critically evaluating the answers. Based on various studies in the field of artificial intelligence, it can be concluded that the performance of AI chatbots may vary across different scientific domains. Overall, in the field of knowledge management, Gemini performed better than the other chatbots evaluated in this research. This study is the first research to examine the performance of AI chatbots in the field of knowledge management. Furthermore, this topic represents a scientific contribution that can be beneficial to AI researchers in improving chatbot capabilities and designing localized chatbots, as well as to knowledge management professionals in effectively utilizing chatbots within organizations.