Recommending valuable ideas in an open innovation community. A text mining approach to information overload problem

Pages683-699
Publication Date14 May 2018
DOIhttps://doi.org/10.1108/IMDS-02-2017-0044
AuthorHanjun Lee,Keunho Choi,Donghee Yoo,Yongmoo Suh,Soowon Lee,Guijia He
SubjectInformation & knowledge management,Information systems,Data management systems,Knowledge management,Knowledge sharing,Management science & operations,Supply chain management,Supply chain information systems,Logistics,Quality management/systems
Recommending valuable ideas in
an open innovation community
A text mining approach to information
overload problem
Hanjun Lee
Korea Institute for Defense Analyses, Seoul, South Korea
Keunho Choi
Department of Business Administration, Hanbat National University, Daejeon,
South Korea
Donghee Yoo
Department of Management Information Systems, BERI,
Gyeongsang National University, Jinju, South Korea
Yongmoo Suh
College of Business Administration, Korea University, Seoul, South Korea
Soowon Lee
School of Software, Soongsil University, Dongjak-gu, South Korea, and
Guijia He
Department of Computer Science and Engineering, Soongsil University,
Seoul, South Korea
Abstract
Purpose Open innovation communities are a growing trend across diverse industries because they provide
opportunities of collaborating with customers and exploiting their knowledge effectively. Although open
innovation communities can be strategic assets that can help firms innovate, firms nonetheless face the
challenge of information overload incurred due to the characteristic of the community. The purpose of this
paper is to mitigate the problem of information overload in an open innovation environment.
Design/methodology/approach This study chose MyStarbucksIdea.com (MSI) as a target open
innovation community in which customers share their ideas. The authors analyzed a large data set collected
from MSI utilizing text mining techniques including TF-IDF and sentiment analysis, while considering both
term and non-term features of the data set. Those features were used to develop classification models to
calculate the adoption probability of each idea.
Findings The results showed that term and non-term features play important roles in predicting the
adoptability of ideas and the best classification accuracy was achieved by the hybrid classification models.
In most cases, the precisions of classification models decreased as the number of recommendations increased,
while the modelsrecalls and F1s increased.
Originality/value This research dealt with the problem of information overload in an open innovation
context. A large amount of customer opinions from an innovation community were examined and a
recommendation system to mitigate the problem was proposed. Using the proposed system, the firm can get
recommendations for ideas that could be valuable for its business innovation in the idea generation phase,
thereby resolving the information overload and enhancing the effectiveness of open innovation.
Keywords Sentiment analysis, Text mining, Open innovation, Data mining, MyStarbucksIdea.com,
Recommendation system
Paper type Research paper
1. Introduction
Innovationis important for every organizationto survive in todays increasingly com plex and
turbulent environment (Chesbrough, 2003; Friedman, 2011; Lee et al., 2012). The need for
innovationhas been a priority of CEOs (Andrew et al., 2010;Jaruzelski and Dehoff, 2010)and a
Industrial Management & Data
Systems
Vol. 118 No. 4, 2018
pp. 683-699
© Emerald PublishingLimited
0263-5577
DOI 10.1108/IMDS-02-2017-0044
Received 3 February 2017
Revised 28 August 2017
Accepted 10 October 2017
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0263-5577.htm
683
Valuable ideas
in an open
innovation
community
popular issue for academic researchers (Hauser et al., 2006; Krishnan and Ulrich, 2001).
Although there isno commonly accepted definition of innovation, it can be defined simply as
the successful exploitation of new ideas to create new business values (Adams et al., 2006;
Francis and Bessant, 2005; Gopalakrishnan and Damanpour, 1997). Thus, the generation of
new ideas andknowledge that meet various customerrequirements plays an important role in
the innovation process (Gloet and Terziovski, 2004). In that sense, ideas are considered the
seeds of innovation and critical determinant of success in managing innovation.
Traditionally, firms have generated such ideas through invention, development, and design,
and have relied on in-house teams of professional inventors whose ideas have led to competitive
advantages. Most firms have taken this closedapproach to innovation; accordingly,
innovations have been restricted by internal resources (Ahlstrom, 2010; March, 1991; Wyld,
2010). Rapid change and intense competition in todays business environment, however, has
altered the way firms initiate and develop innovations. Since it is almost impossible for firms to
obtain all the resources they require for innovation, they have begun to acquire external
knowledge to complement their limited internal capabilities (Beamish and Lupton, 2009;
Cassiman and Veugelers, 2006). Consequently, innovation-related activities such as concept
generation, needs-finding, and idea generation, which have traditionally been done using a top-
down approach, are now done using a bottom-up approach that includes customers because
firms regard this as more effective. In light of this trend, the concept of open innovation
represents a paradigm that emphasizes the use of knowledge from outside an organization as
well as inside (Chesbrough, 2003). This paradigm advocates that organizations should use more
open strategies to complement internal innovation processes (Chesbrough et al.,2008;Laursen
and Salter, 2005). Accordingly, firms from different industries, such as Dell, Starbucks, and P&G,
have adopted open strategies by managing their own online open innovation communities. Such
communities can be strategic assets for organizational innovation by: providing external
expertise, generating valuable ideas, and supporting innovation development (Dahlander and
Wallin, 2006). Starbucks, for instance, gathers more than 1,500 ideas per week, and has
accumulated more than 150,000 from its online open innovation community to date. These ideas
are discussed, developed, and adopted to help Starbucks maintain its innovation strategy.
In most communities, the more ideas there are the better. A large number of ideas
increase diversity and are likely to result in more innovative options for firms. However, this
situation can also create a serious problem of information overload. As discussed above,
open innovation necessarily requires a discussion about each idea; consequently, a large
number of ideas make the process difficult and ineffective. Active idea generation can
therefore hinder effective idea selection. The objective of this paper is to mitigate the
problem of information overload in an open innovation context by recommending the top
nideas with the highest adoption probabilities. To this end, we first analyzed the data set
collected from MyStarbucksIdea.com (MSI), utilizing text mining techniques that included
term frequency-inverse document frequency (TF-IDF) and sentiment analyses. We then
used the results from the analyses as input variables to build a system using data mining
techniques that recommends innovative ideas to Starbucks.
By using such a recommendation system, it is possible for firms to receive
recommendations about those ideas that could be useful for innovation at the initial
ideation stage. This can enhance effectiveness and efficiency for firms when they process
customersideas obtained from open innovation communities.
The rest of this paper is organized as follows. In the next section, we review previous
research on open innovation and sentiment analysis. Section 3 presents our proposed
method, including the research framework and model building. Section 4 presents an
experimental design, including data set, preprocessing, and feature selection. In Section 5,
we describe the experiments results. In the last Section, we discuss implications of our
results and then we conclude.
684
IMDS
118,4

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT