Mapping Development of Open Innovation Visually and Quantitatively: A Method of Bibliometrics Analysis

Open innovation has become one of the hottest fields in innovation management. Although some reviews have been done to conclude the work of open innovation study, however, few of them were conducted quantitatively as well as visually. From a perspective of bibliometrics analysis, this article depicted a map of development of open innovation study with the help of CiteSpace II. Most important references and authors about open innovation study were picked out objectively according to their ranks and the visualization. The process of open innovation development was described and the newest topics in recent years are highlighted to help understand history of open innovation and do further study in this field. Also, according to the trend study, the current stage of open innovation study was estimated and the trends of recent future were forecasted.

Although open innovation obtains some agreement among researchers, it is still a relative new field. And most of the studies focus on its nature rather than its powerful application in industries. Some work should be done to track trend of open innovation study to provide information about its history and future development. This paper mapped the history of open innovation study, found the hot topics in recent years and depicted the trend of open innovation study visually and quantitatively using bibliometrics analysis method with the help of Web of Science and CiteSpace II (Chen, 2004(Chen, , 2006.

Methodology
This study used data got from Web of Science and CiteSpace II to analyze the trend of open innovation in the history and in the future. Although Web of Science only contains part of all the papers about open innovation, their references cover almost all the ones in this area. In this case, we can depict a map of the evolution of open innovation. In order to make sure papers collected from Web of Science can reflect the whole trend, Web of Knowledge which includes more ones was taken in to test it. We searched for all the articles and reviews in all years with a term "open innovation". Finally, we found 1564 records in Web of Science and 6028 ones in Web of Knowledge. Although they have quite different number, the number of cited papers is relatively the same. For the bibliometrics analysis, the useful ones are the top ones which indicates that papers that are out of Web of Science can hardly make negative effect on our analysis. At the same time, the trends of both search results are quite alike (see figure 1 to 4).

Important References in Open Innovation Study
Firstly, references of papers in the dataset were analyzed by CiteSpace II, top 1% papers every year in the dataset were selected out for analysis. With the help of CiteSpace II, the top 5 cited references can be found according to their frequency of being cited. The book "Open Innovation: The new imperative for creating and profiting from technology" written by Henry Chesbrough in 2003 is the most important one now. In this book, Henry Chesbrough firstly proposed the item "Open innovation" formally and gave out cases indicating the power of open innovation. The existence of the second one "Absorptive Capacity: A New Perspective on Learning and Innovation" supports the idea that study and practice of open innovation were common in the past even though the item did not appear and studies and practices were not systematic.   At last, in order to see the map more clearly, we chose to work on top 10 cited references for each slice (figure 9) which enjoys the same structure as figure 6.

Main Authors for Open Innovation Study
We list the top 10 authors according to both their numbers in the dataset (Table 2) with help from Web of Science and how many times they were cited (Table 3) with help from CiteSpace II. In fact, "Chesbrough H.W." and "Chesbrough H" in Table 3 Figure 10).

The Trend of Open Innovation Study in the Past and in the Recent Future
Firstly, papers in our dataset were studied according to the Web of Science Categories (Table 4). From this result, we believe that thoughts of open innovation have been applied to many fields and paly outstanding role in certain ones (like "computer science information systems" and "surgery"). "Management" ranks firstly in this www.ccsenet.org/ass Asian Social Science Vol. 9, No. 11;2013 dataset which indicates that general study on open innovation is still prevalent. In this case, open innovation cannot be treated as a mature discipline. However, the ones rank after "Management" are "Engineering industrial", "Operations research management science", "Computer science information systems", "Surgery" and so on. They take much more percentages than "Management" in this dataset which reveals that open innovation has already been used to some relative fields. From this situation, we can know that open innovation obtains some useful theories and accumulates some knowledge but still needs further study to answer most questions about it. Besides the dataset, we mostly focused on cited references to depict trends of open innovation. Key words of cited papers were studied with CiteSpece II. Firstly, top 10% and used 1-year slices was chosen and threshold was made be 100. Based on this, we listed top 20 cited keywords in period 1990 to 2013 (Table 5) and in period 2010 to 2013 (Table 6). Combine the two lists, we can easily find that the key word "research-and-development" exceeded "management" and "model" and ranked 5 which shows it became more important these years than before. While "model", "open source software", "evolution" and "diffusion" dropped down significantly. This phenomenon indicates study of open innovation these days less focus on model study, knowledge diffusion and evolution and software study no longer plays that important role in open innovation. Furthermore, "information" and "surgery" are not in the top 20 key words any more. And "absorptive ability" and "network" took their place. This result means that open innovation study is becoming more and more concentrated on its own area instead of surgery.   Figure 11. Change of key words in rank Figure 12. Change of key words in centrality According to figure 13 all the keywords about open innovation linked tightly which also means this is a relative new field and all topics have some relationship with each other. According to the timeline mode of these key words, open innovation developed from a distinct road ( Figure 14). Firstly, "model", "diffusion" and "innovation" provides basic ideas for development of open innovation. After that, the perspective of system plays an important role. Then it comes to "internet" and "management". People became focus on technology, information and performance of organizations. "Research and development" is also an important topic for open innovation study. Then with all these keywords, open innovation was formally proposed in 2003. After this item, researchers study further on "evolution", "knowledge", "design", "open source software", "perspective", "industry", "product development", "impact", "absorptive ability" and "network".
In order to see structure of these keywords more clearly, we marked clusters worked out by CiteSpace II ( Figure  15) and reset to pick out top 30 ( Figure 16  , "open innovation", "performance", "technology", "research-and-development", "model", "management", "industry", "product development", "systems", "open source software", "design", knowledge", "absorptive-capacity", "firms", "perspective", " impact", "networks", "evolution" and "diffusion". 5) The hot and new topics were found and topics becoming less attractive were also picked out. Among all the top 20 key words, "research-and-development" attracted more attentions while "model", "open source software", "evolution" and "diffusion" dropped down significantly. "Absorptive capacity" and "network" are new ones in this family. And "information" and "surgery" were no longer ranked in top 20. In one word, Researchers pay attention to Research and Development, the ability of absorbing knowledge and networks of firms much more than before.
6) Open innovation was formed directly along a clear road. Figure 14 shows its road map. We can list all the points on this road in a time order: firstly it was "model", "diffusion" and "innovation"; then "internet" and "management"; "technology", "information" and "performance"; "research and development". And with all these keywords, "open innovation" was formally proposed in 2003. After this item, researchers study further on "evolution", "knowledge", "design", "open source software", "perspective", "industry", "product development", "impact", "absorptive ability" and "network".

7)
Open innovation is at its early stage. Open innovation study attracts many researchers and organizations. However, it has not systematic theories or knowledge although it really helped some companies to get high profits. At this stage, most studies are still focus on the nature of open innovation itself rather than develop its applicable branches. It must need some more time to find or develop its useful branches for the further contribution to problems in reality.

Conclusion
This Also, there are some shortcomings in our study. Firstly, the dataset is a misleading one to some extent. Many important journals are excluded. Even though their cited references can describe situations quite well, we lose the opportunity to process target data directly. Besides, figures made by CiteSpace II cannot contain all data because of its computational capability. In this case, minor details will be ignored.
Open innovation is a hot topic nowadays. After Henry Chesbrough's great work, many companies in almost all industries are interested in this fantastic idea about technology management. Open innovation does provide a good way for future development of many organizations. In this condition, it is becoming extremely hot. However, open innovation is at an early stage according to our study. Its nature and application need powerful study to contribute to future innovation and technology management. Based on our work, studying at nature of open innovation will still be the most important task in recent future but finding or developing new branches from open innovation will be an important and long term trend.