Detecting Feature Requests of Third-Party Developers through Machine Learning: A Case Study of the SAP Community - Proceedings of the 56th Hawaii International Conference on System Sciences

Aug 4, 2026

Upload by Jiven

Requirements elicitation is fundamental for successful software product development, yet traditional methods like interviews are labor-intensive and difficult to scale. For enterprise vendors, eliciting requirements from users is especially challenging because app store reviews and direct user access are limited. This study explores whether enterprise vendors can mine requirements from their sponsored developer communities. Using data from the SAP Community and a supervised machine learning classifier trained on 1,500 manually labeled questions, the approach achieves accuracy of 0.819 and effectively identifies feature requests.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Abstract

1. Introduction

Requirements elicitation: definitions and challenges

Data-driven requirements elicitation and supporting technologies

Why enterprise software requirements elicitation remains difficult

Proceedings of the 56th Hawaii International Conference on System Sciences | 2023 Detecting Feature Requests of Third-Party Developers through Machine Learning: A Case Study of the SAP Community Martin Kauschinger Technical University of Munich [martin.kauschinger@tum.de](martin.kauschinger@tum.de) Niklas Vieth SAP Deutschland SE & Co. KG [niklas.vieth@sap.com](niklas.vieth@sap.com) Maximilian Schreieck University of Innsbruck [maximilian.schreieck@uibk.ac.at](maximilian.schreieck@uibk.ac.at) Abstract The elicitation of requirements is central for the development of successful software products. While traditional requirement elicitation techniques such as user interviews are highly labor-intensive, data-driven elicitation techniques promise enhanced scalability through the exploitation of new data sources like app store reviews or social media posts. For enterprise software vendors, requirements elicitation remains challenging because app store reviews are scarce and vendors have no direct access to users. Against this background, we investigate whether enterprise software vendors can elicit requirements from their sponsored developer communities through data-driven techniques. Following the design science methodology, we collected data from the SAP Community and developed a supervised machine learning classifier, which automatically detects feature requests of third-party developers. Based on a manually labeled data set of 1,500 questions, our classifier reached a high accuracy of 0.819. Our findings reveal that supervised machine learning models are an effective means for the identification of feature requests. Keywords: Data-Driven Requirements Engineering, Enterprise Software, Machine Learning, Online Community, Platform Ecosystem 1. Introduction The accurate elicitation of requirements is central for the development of successful software products (Chakraborty et al., 2010; Meth et al., 2015) . Unfortunately, both researchers and practitioners have observed that many software development projects fail due to inaccurate or missing user requirements, which often result in significant financial losses for the development firm (Mathiassen et al., 2007; Rosenkranz et al., 2014) . Specifically, the communication and interactions between the various stakeholders make the requirements elicitation process complex and hard to manage. Moreover, several studies have shown that users often lack the ability to specify their requirements Helmut Krcmar Technical University of Munich [helmut.krcmar@tum.de](helmut.krcmar@tum.de) correctly (Hansen & Lyytinen, 2010; Rosenkranz et al., 2014) . Requirements elicitation refers to the identification and extraction of conscious, unconscious and subconscious requirements of all involved stakeholders of a development project (Saiediana & Daleb, 2000) . Although requirements elicitation has been an important topic in the information systems field for several decades (Chakraborty et al., 2010; Rosenkranz et al., 2014), its techniques have changed substantially in recent years. Previously, requirements elicitation was largely based on traditional elicitation techniques such as on-site observations, user interviews, focus groups or workshops (Byrd et al., 1992; Saiediana & Daleb, 2000) . These traditional techniques require close and frequent interactions between system analysts and users, making them highly labor- and cost-intensive (Chakraborty et al., 2010) . More recently, researchers and practitioners have acknowledged the rise of datadriven requirements elicitation techniques (Maalej et al., 2016; Meth et al., 2015) . These data-driven techniques integrate newly available data sources such as app store reviews or social media posts into the elicitation process (e.g., Halckenhaeußer et al., 2022; Hoffmann et al., 2019; Kauschinger et al., 2021; Maalej & Nabil, 2015) . Moreover, data-driven techniques often rely on new analytical technologies such as natural lang

FAQs

Why is requirements elicitation central to software product success?

Accurate elicitation helps ensure that development teams capture the needs of involved stakeholders. Missing or inaccurate requirements can cause significant project failure and financial loss.

What challenge do enterprise vendors face when using data-driven elicitation techniques?

Enterprise vendors lack direct access to end users, and app store reviews are scarce for enterprise software, limiting the availability of suitable data sources.

How does the study detect feature requests from third-party developers?

The study follows a design science methodology: it collects SAP Community data and trains a supervised machine learning classifier on 1,500 manually labeled questions to identify feature requests automatically.

Aug 4, 2026

Upload by Jiven