دانلود مقاله طراحی عامل مکالمه ای برای تسهیل ایده پردازی
چکیده
مقدمه
کار مرتبط
رویکرد تحقیق
اهداف یک راه حل
طراحی و توسعه
تظاهرات
ارزیابی
بحث
نتیجه
منابع
Abstract
Introduction
Related Work
Research Approach
Objectives of a Solution
Design and Development
Demonstration
Evaluation
Discussion
Conclusion
References
چکیده
تعداد زیادی از ارسال‌های ناقص، نامشخص و نامشخص در پلت‌فرم‌های ایده، سازمان‌ها را برای بهره‌برداری از پتانسیل کامل طرح‌های نوآوری باز باز می‌دارد، زیرا انتخاب ایده دست‌وپاگیر است. در یک پروژه تحقیقاتی علم طراحی، ما طرحی را برای یک عامل مکالمه (CA) مبتنی بر هوش مصنوعی ایجاد می‌کنیم تا مشارکت‌کنندگان را در ایجاد ایده‌های پیچیده در پلت‌فرم‌های ایده‌ای که در آن تسهیلات انسانی مقیاس‌پذیر نیست، تسهیل کنیم. ما دانش طراحی تجویزی را در قالب اصول طراحی استخراج می‌کنیم، CA را در دو قسمت ارزیابی متوالی مورد بررسی قرار می‌دهیم و ارزیابی می‌کنیم. اصول طراحی به جریان تحقیق فعلی در زمینه تسهیل خودکار کمک می کند و می تواند ارائه دهندگان پلتفرم های ایده را برای بهبود تولید ایده و فرآیندهای انتخاب ایده بعدی راهنمایی کند. نتایج نشان می‌دهد که تسهیلات مبتنی بر CA برای مشارکت‌کنندگان جذاب است و ایده‌های ساختاریافته و دقیقی را ارائه می‌دهد.
توجه! این متن ترجمه ماشینی بوده و توسط مترجمین ای ترجمه، ترجمه نشده است.
Abstract
Large numbers of incomplete, unclear, and unspecific submissions on idea platforms hinder organizations to exploit the full potential of open innovation initiatives as idea selection is cumbersome. In a design science research project, we develop a design for a conversational agent (CA) based on artificial intelligence to facilitate contributors in generating elaborate ideas on idea platforms where human facilitation is not scalable. We derive prescriptive design knowledge in the form of design principles, instantiate, and evaluate the CA in two successive evaluation episodes. The design principles contribute to the current research stream on automated facilitation and can guide providers of idea platforms to enhance idea generation and subsequent idea selection processes. Results indicate that CA-based facilitation is engaging for contributors and yields well-structured and elaborated ideas.
Introduction
Organizations face challenges in discovering and developing innovations due to limited internal resources (Hansen & Pries-Heje, 2017) and the fact that “when focusing on a limited solution space, companies only apply the most obvious instead of the most efficient of all solutions in order to solve an innovation problem” (Lüttgens et al, 2014, p. 342). In this regard, open innovation approaches have been identified to be an effective strategy to improve the efficacy of organizations’ innovation capabilities (Chesbrough, 2003; Lüttgens et al., 2014). Digital platforms, e.g. idea platforms, enable organizations to apply idea sourcing by involving external contributors to access widely dispersed external knowledge and expertise beyond their boundaries (Boudreau & Lakhani, 2013; Cricelli et al., 2021; Di Gangi & Wasko, 2009). However, organizations struggle to harness the potential of idea platforms (Piezunka & Dahlander, 2015), as such idea sourcing initiatives generate highly diverse input whose utilization and valorization remains a key challenge. In particular, the large quantity of contributions pose major challenges in terms of textually unstructured ideas with an insufficient level of detail and indistinct causalities (Barbier et al., 2012; Kipp et al., 2013). As a result, organizations invest a great expenditure of human capacity and time during idea selection to organize and evaluate ideas to select those with high potential (Blohm et al., 2013; Kittur et al., 2013; Merz, 2018). Nevertheless, familiar contributions or ideas with detailed information but little implementation potential might be selected over those with a lack of details and great potential (Bansemir & Neyer, 2009; Piezunka & Dahlander, 2015).
Conclusion
As part of a multi-cycle DSR research project, this study presents a solution to elevate organizational idea generation processes on idea platforms with AI-based CA technology. While idea generation facilitation is critical to innovation, organizations struggle to leverage this potential on idea platforms. So far, large amounts of ambiguous, imprecise, and incomplete ideas hamper organizations in selecting ideas with potential for further processing. To address these challenges, we built on the facilitation concept to iteratively design and instantiate a scalable CA that facilitates individuals during their idea generation. Evaluation results suggest that the natural, dialog-based interaction encourages and engages idea contributors to provide clear, detailed, and complete ideas, which deliver a suitable grounding for the essential follow-up selection of textual ideas in organizations.