Quotation Kretzschmar, Linn. 2019. Machine learning-based text mining for evaluating crowdsorcing contributions. Open and User Innovation Conference (OUI), Utrecht, Niederlande, 08.07.-10.07.


RIS


BibTeX

Abstract

Organizations increasingly tap into external sources such as crowds in order to find innovative input (Hansen and Birkinshaw, 2007). Many of these crowdsourcing activities take place online and oftentimes can result in large amounts of data that are characterized by redundancy and high variationin quality (Toubia and Florès, 2007). Once an organization has acquired this large amount of data, one of the biggest challenges is to identify the « best » ideas for its purposes. This step is known to be a big bottleneck in the innovation process since it is cost-intensive and requires temporal commitment (Rietzschel et al., 2010). While the standard approach to evaluating ideas is either employing expert raters or harnessing crowds, these methods are particularly ineffective when it comes to the evaluation of a large data corpus that contains idea descriptions provided as unstructured texts (Klein and Garcia, 2015). Explanations for the low selection performance of human raters include cognitive overload due to the complexity of the decision-making process and information overload, as well as cognitive biases, such as familiarity bias (Westerski et al., 2013; Rietzschel et al., 2010; Brennan et al., 2010). Therefore, some researchers suggest utilizing algorithmic approaches, which have become more popular since the rise of machine learning (ML) and that have proven to be successfully applicable in various text-mining contexts (Toubia and Florès, 2007, Klein and Garcia, 2015). As a subset of artificial intelligence, machine learning enables computer systems to apply algorithms and statistical models that automatically learn and improve with experience and over time. In the business context, organizations already take advantage of these technological advances in areas like analyzing customer feedback and customer sentiments, predicting user behavior and generating product recommendations. In recent years, some researchers have shown that machine learning approaches [...]

Tags

Press 'enter' for creating the tag

Publication's profile

Status of publication Published
Affiliation WU
Type of publication Paper presented at an academic conference or symposium
Language English
Title Machine learning-based text mining for evaluating crowdsorcing contributions
Event Open and User Innovation Conference (OUI)
Year 2019
Date 08.07.-10.07.
Country Netherlands
Location Utrecht
URL https://sites.google.com/view/oui2019/program-oui-2019-at-utrecht-university/daily-schedule-oui-2019

Associations

People
Kretzschmar, Linn (Former researcher)
Organization
Institute for Entrepreneurship and Innovation IN (Details)
Google Scholar: Search