狠狠撸

狠狠撸Share a Scribd company logo

Sasaki.informs2014(2)

?Download as PPTX, PDF?
?
Hajime Sasaki
Hajime Sasaki

This document proposes a methodology to detect knowledge transfer from science to technology using network analysis and bibliometrics. As a case study, it analyzes the fields of academic papers and patented technologies in solar cells. It clusters papers and patents by citation networks and identifies characteristic keywords in each cluster. It then calculates semantic similarities between clusters to create a meta-network and detects the maximum flow route, identifying a boundary cluster related to "dye sensitized" research that transferred from academic papers to patented technologies.

1 of 19
Detecting Knowledge Transitions Between Science and Technology for Forecasting Growing Fields. 1 Hajime SASAKI*1 Yuya KAJIKAWA*2 Ichiro SAKATA*1 *1)Innovation Policy Research Center, The University of Tokyo. and Policy Alternatives Research Institute, The University of Tokyo. *2)Innovation Management, Tokyo Institute of Technology.
Background (1): Relationships between Science and Technology “Close links between academia and industry have many positive aspects not only for the business partner(Zucker and Darby, 2000; Hall et al., 2001) but also for the academic sector. ” (Dirk et al., 2009) “Industry–science collaborations might even trigger new basic research.” (Rosenberg, 1998) “A positive relationship of patenting activities and publication outcome as well as publication quality.” (VanLooy et al., 2006; Czarnitzki et al., 2007; Breschi et al., 2007; Azoulay et al., 2006, etc.).
Background(2): Fundamental Challenges We Face ? Exponential growth of knowledge We are drowning in the information sea. Volume of Knowledge Industry Revolution Time Renaissance Before modern history Prerequisite to catch up with the pace of development Time ? Segmentation & Specialization Difficult to grasp the whole picture in each field even specialists 3
Research Question Question ? Can we detect a transition of knowledge from science to technology and contribute to an early detection of promising technologies in advance? Purpose ? To propose a methodology to detect knowledge transfer from science to technology, which will become growing field using network analysis and bibliometrics.
Citation Network Analysis ? Node： Papers or Patents in the largest-graph component in network data ? Link： Citation between Papers( or Patents) ? Year： Average Publication Year in the cluster ? Keywords: Characteristic words in the cluster Node (Paper) Link (Citation) Cluster 3 Node: 6 Edge: 6 Year: 2007.96 Keywords: ---- Cluster 1 Node: 5 Edge: 4 Year: 2003.21 Keywords: ---- Cluster 2 Node: 3 Edge: 3 Year: 2004.36 Keywords: ---- 5 Example)
Methodology Flow Science Layer Technology Layer Extract Patent Dataset Common process Create Citation Network (by year) Network Clustering Keyword Recognition Determine Similarity Create Meta Network Detect max flow route Extract Paper Dataset 1 2 3 4 5 6 Node: Paper or Patent Link: Citation Node: Cluster Link: Similarity
Dataset: Solar cell (Photovoltaic, PV) 7 Science Layer Dataset: Academic Paper Database: Thomson Web of science? Queries“photovoltaic” OR “solar cell” Number of Papers: 50,913 Technology Layer Dataset: Patent Gazzet Database: Thomson Innovation? Queries“photovoltaic” OR “solar cell” Number of Patents: 63,972
Network Clustering (Newman M.E.J, 2004) wij: the possibility of the weights of edges in the network that connected nodes in cluster s to those in cluster j ? Connect clusters sparsely and extract clusters within which nodes are connected densely is cut. ? A high value of Q represents good community division where only dense edges remain within clusters and sparse edges between clusters are cut off, and ? Q = 0 means that a particular division gives no more within-community edges. 8
Characteristic Keywords in each Cluster: (tf-idf) 9 ? TFIDF – Good for larger tf (Ferm Frequency) – Good for small df (Document Frequency) tfidf(d, w) = tf(d,w) idf(w) = tf(d,w) log(N /df(t)) N: number of documents Specific terms
C1(Cluster1) “thin film” C2(Cluster2) “organic” C3(Cluster3) “dye sensitized” C4(Cluster4) “power tracking” C5(Cluster5) Citation Network in Academic Research Dataset: “Solar Cell” published Until 2012 Paper: Node “HgCdTe photovoltaic” Citation: Link Mercury Cadmium Telluride
C1(Cluster1) “conductor” C2(Cluster2) “protection sheet” C3(Cluster3) “dye sensitized” C4(Cluster4) “multi junction” C5(Cluster5) “silicon semiconductor” Citation Network in Patented Technology Dataset: “Solar Cell” published Until 2012
Top 10 keywords in each Cluster Science(Academic Paper) Technology (Patent) C1 untill 2012 C2 untill 2012 C3 untill 2012 C4 untill 2012 C5 untill 2012 film polymer tio2 power hgcdte silicon conjugated dye system detector thin film blend sensitized inverter infrared thin organic sensitized solar module photovoltaic detector deposition p3ht sensitized solar cgerildl photodiodes solar cell poly dye sensitized renewable fesi2 solar pcbm dye sensitized sotrlacrking wavelength layer conjugated polymdyeer sensitized soblaatrt ecreyll array efficiency fullerene electrolyte maximum powerhg1 cdte bulk heterojunctzionno wind focal plane C1 untill 2011 C2 untill 2011 C3 untill 2011 C4 untill 2011 C5 untill 2011 film polymer tio2 power hgcdte silicon conjugated dye system detector thin film p3ht sensitized module infrared thin blend dye sensitized inverter photovoltaic detector deposition organic sensitized solar tracking fesi2 solar cell pcbm sensitized solar cgerildl photodiodes solar poly dye sensitized sorelanrewable wavelength layer conjugated polymdyeer sensitized sowlainr dcell hg1 efficiency bulk heterojuncteiolenctrolyte energy array cdte fullerene zno battery plane array 12 C1 untill 2012 C2 untill 2012 C3 untill 2012 C4 untill 2012 C5 untill 2012 photovoltaic resin dye layer type electrode sheet sensitized oxide electrode contact polyester dye sensitized silicon silicon tab sealing sensitized solar subcell semiconductor substrate layer protection sheetdye sensitized sojulanrction etching module copolymer sensitized solar ctrealnl sparent condwucirtiinvge photovoltaic cellsolar cell moduldeye sensitized socloanr dcueclltive surface portion sheet for solar ceelllectrolyte transparent portion material sheet for solar electrode semiconductor substrate conductive cell module sensitizing solar subcell impurity C1 untill 2011 C2 untill 2011 C3 untill 2011 C4 untill 2011 C5 untill 2011 layer resin dye transparent organic photovoltaic sheet sensitized film dye electrode polyester dye sensitized oxide photoactive module solar cell modulesensitized solar layer electrode contact sealing dye sensitized soslialircon layer photovoltaic cellcell module sensitized solar ceelellctrode material forming sheet for solar cedlyle sensitized sotrlanr scpeallrent condeulcetcitvreon subcell sheet for solar electrode conductive comprises surface copolymer sensitizing amorphous photovoltaic tab module electrolyte plasma sensitized Until 2010 ? Until 2009 ? ? Until 2010 ? Until 2009 ? ? Until 2012 Until 2011 Until 2012 Until 2011
Semantic Similarites between Clusters 0.6 0.5 0.4 0.3 0.2 0.1 0 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 Patent Academic Paper C1 C2 C3 C4 C5 Academic Paper until 2012 C1 C2 C3 C4 C5 Patent until 2012 Legend 0 0.6 Cosine Similarity
Maximum Flow in Network 14 Node: Cluster Link: Semantic Similarity t t 0 t 1 t 1 +1 t end Meta Network Example) Source Sink
Technology Layer (Patented Technologies) 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Science Layer (Academic Papers) Source Sink C5 C4 C3 C2 C1 C1 C2 C3 C4 C5
Technology Layer (Patented Technologies) 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 C5 C1 C2 Source Sink C4 C3 C2 C1 C3 C4 C5 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Science Layer (Academic Papers)
17 Top 10 Keywords in each boundary cluster Science Layer Technology Layer Cluster 2. in Academic Paper 学until 術2005 論文(2005. C2) 特許公報(2007. C2) tio2 dye dye sensitized polymer dye sensitized sensitized dye sensitized solar dye sensitized dye sensitized solar cell sensitized solar sensitized solar cell sensitized solar cell sensitized solar nanocrystalline sensitizing dye sensitized solar pigment dye sensitized solar cell electrode Cluster 2. in Patent until 2007
Summary ? It is necessary to understand knowledge relationships between Science and Technology, for Innovation Strategy. ? We considered the relationship as a time expanded and Heterogeneous network. ? We also considered this problem can be interpreted as a Maximum flow problem of Semantic Similarity Network. ? Photovoltaic field as case study. ? Grasped boundary clusters related to “Dye sensitized” field. 18
Thank you for your attention. References ? Zucker, L.G., Darby, M.R.,“Capturing technological opportunities via Japan’s star scientists.”, Journal of Technology Transfer 26, 37–58, 2000. ? Hall B.H., Link A.N., Scott J.T., “Barriers inhibiting industry from partnering with universities: evidence from the advanced technology program.”, Journal of Technology Transfer 26, 87–98, 2001. ? Azoulay P., Ding W., Stuart T., 2006. “The impact of academic patenting on the rate, quality and direction of (public) research.”, NBER working paper 11917, Cambridge, MA., 2006. ? Rosenberg N., “Chemical engineering as a general purpose technology. In: Helpman, E. (Ed.), General Purpose Technologies and Economic Growth.”, MIT Press, Cambridge, pp. 167–192, 1998. ? Francis Narin, Kimberly S. Hamilton and Dominic Olivastro, “The increasing linkage between U.S. technology and public science”, Research Policy, Volume 26, Issue 3, October 1997, Pages 317–330, 1997. ? N. Shibata, Y. Kajikawa, and I. Sakata. “Extracting the commercialization gap between science and technology - case study of a solar cell.”, Technological Forecasting and Social Change, 77:1147–1155, 2010. ? Small H., “Citation Structure of an Emerging Research Area: Organic Thin Film.”, Proceedings of ISSI, pp. 718-725. 2007. ? 七丈, “共引用クラスタリングによる研究分野の動的把握に向けた試論”, 情報知識学会誌2013, Vol.23, No.3, 371-379, 2013. ? F.G. Engineer, G.L. Nemhauser, and M.W.P. Savelsgergh, “Dynamic programming-based column generation on Time-Expanded Network: Application to the Dial-a-Flight problem”, INFORMS Journal on Computing, 23(1), pp. 105-119, 2011. ? N. Shah, S. Kumar, F. Bastani, and I.L. Yen, “Optimization models for assessing the peak capacity utilization of intelligent transportation systems”, European Journal of Operational Research, 216, pp. 239-251, 2012. ? Newman M.E.J., “Fast algorithm for detecting community structure in networks”, Physical Review E, Vol. 69, p. 066133, 2004. ? Ino, H, Kudo. M, Nakamura. A, “A Comparative Study of Algorithms for Finding Web Communities”, Data Engineering Workshops, 2005. 21st International Conference, 1257, 2005. ? Horiike. T, Takahashi. Y, Kuboyama. T and Sakamoto. H, “Extracting Research Communities by Improved Maximum Flow Algorithm”, Knowledge-Based and Intelligent Information and Engineering Systems Lecture Notes in Computer Science Volume 5712, pp472-479, 2009. ? Victor Str?ele, Geraldo Zimbr?o, Jano M. Souza, “Modeling, Mining and Analysis of Multi-Relational Scientific Social Network”, Journal of Universal Computer Science, vol. 18, no. 8 (2012), 1048-1068, 2012. ? A. V. Goldberg and R. E. Tarjan, “A New Approach to the Maximum Flow Problem”, Journal of the ACM 35:921-940, 1988. ? Péter ?rdi, Kinga Makovi, Zoltán Somogyvári, Katherine Strandburg, Jan Tobochnik, Péter Volf, László Zalányi, “Prediction of Emerging Technologies Based on Analysis of the U.S. Patent Citation Network”, Scientometrics: Volume 95, Issue 1 (2013), Page 225-242, 2013.

More Related Content

Sasaki.informs2014(2)

  • 1. Detecting Knowledge Transitions Between Science and Technology for Forecasting Growing Fields. 1 Hajime SASAKI*1 Yuya KAJIKAWA*2 Ichiro SAKATA*1 *1)Innovation Policy Research Center, The University of Tokyo. and Policy Alternatives Research Institute, The University of Tokyo. *2)Innovation Management, Tokyo Institute of Technology.
  • 2. Background (1): Relationships between Science and Technology “Close links between academia and industry have many positive aspects not only for the business partner(Zucker and Darby, 2000; Hall et al., 2001) but also for the academic sector. ” (Dirk et al., 2009) “Industry–science collaborations might even trigger new basic research.” (Rosenberg, 1998) “A positive relationship of patenting activities and publication outcome as well as publication quality.” (VanLooy et al., 2006; Czarnitzki et al., 2007; Breschi et al., 2007; Azoulay et al., 2006, etc.).
  • 3. Background(2): Fundamental Challenges We Face ? Exponential growth of knowledge We are drowning in the information sea. Volume of Knowledge Industry Revolution Time Renaissance Before modern history Prerequisite to catch up with the pace of development Time ? Segmentation & Specialization Difficult to grasp the whole picture in each field even specialists 3
  • 4. Research Question Question ? Can we detect a transition of knowledge from science to technology and contribute to an early detection of promising technologies in advance? Purpose ? To propose a methodology to detect knowledge transfer from science to technology, which will become growing field using network analysis and bibliometrics.
  • 5. Citation Network Analysis ? Node： Papers or Patents in the largest-graph component in network data ? Link： Citation between Papers( or Patents) ? Year： Average Publication Year in the cluster ? Keywords: Characteristic words in the cluster Node (Paper) Link (Citation) Cluster 3 Node: 6 Edge: 6 Year: 2007.96 Keywords: ---- Cluster 1 Node: 5 Edge: 4 Year: 2003.21 Keywords: ---- Cluster 2 Node: 3 Edge: 3 Year: 2004.36 Keywords: ---- 5 Example)
  • 6. Methodology Flow Science Layer Technology Layer Extract Patent Dataset Common process Create Citation Network (by year) Network Clustering Keyword Recognition Determine Similarity Create Meta Network Detect max flow route Extract Paper Dataset 1 2 3 4 5 6 Node: Paper or Patent Link: Citation Node: Cluster Link: Similarity
  • 7. Dataset: Solar cell (Photovoltaic, PV) 7 Science Layer Dataset: Academic Paper Database: Thomson Web of science? Queries“photovoltaic” OR “solar cell” Number of Papers: 50,913 Technology Layer Dataset: Patent Gazzet Database: Thomson Innovation? Queries“photovoltaic” OR “solar cell” Number of Patents: 63,972
  • 8. Network Clustering (Newman M.E.J, 2004) wij: the possibility of the weights of edges in the network that connected nodes in cluster s to those in cluster j ? Connect clusters sparsely and extract clusters within which nodes are connected densely is cut. ? A high value of Q represents good community division where only dense edges remain within clusters and sparse edges between clusters are cut off, and ? Q = 0 means that a particular division gives no more within-community edges. 8
  • 9. Characteristic Keywords in each Cluster: (tf-idf) 9 ? TFIDF – Good for larger tf (Ferm Frequency) – Good for small df (Document Frequency) tfidf(d, w) = tf(d,w) idf(w) = tf(d,w) log(N /df(t)) N: number of documents Specific terms
  • 10. C1(Cluster1) “thin film” C2(Cluster2) “organic” C3(Cluster3) “dye sensitized” C4(Cluster4) “power tracking” C5(Cluster5) Citation Network in Academic Research Dataset: “Solar Cell” published Until 2012 Paper: Node “HgCdTe photovoltaic” Citation: Link Mercury Cadmium Telluride
  • 11. C1(Cluster1) “conductor” C2(Cluster2) “protection sheet” C3(Cluster3) “dye sensitized” C4(Cluster4) “multi junction” C5(Cluster5) “silicon semiconductor” Citation Network in Patented Technology Dataset: “Solar Cell” published Until 2012
  • 12. Top 10 keywords in each Cluster Science(Academic Paper) Technology (Patent) C1 untill 2012 C2 untill 2012 C3 untill 2012 C4 untill 2012 C5 untill 2012 film polymer tio2 power hgcdte silicon conjugated dye system detector thin film blend sensitized inverter infrared thin organic sensitized solar module photovoltaic detector deposition p3ht sensitized solar cgerildl photodiodes solar cell poly dye sensitized renewable fesi2 solar pcbm dye sensitized sotrlacrking wavelength layer conjugated polymdyeer sensitized soblaatrt ecreyll array efficiency fullerene electrolyte maximum powerhg1 cdte bulk heterojunctzionno wind focal plane C1 untill 2011 C2 untill 2011 C3 untill 2011 C4 untill 2011 C5 untill 2011 film polymer tio2 power hgcdte silicon conjugated dye system detector thin film p3ht sensitized module infrared thin blend dye sensitized inverter photovoltaic detector deposition organic sensitized solar tracking fesi2 solar cell pcbm sensitized solar cgerildl photodiodes solar poly dye sensitized sorelanrewable wavelength layer conjugated polymdyeer sensitized sowlainr dcell hg1 efficiency bulk heterojuncteiolenctrolyte energy array cdte fullerene zno battery plane array 12 C1 untill 2012 C2 untill 2012 C3 untill 2012 C4 untill 2012 C5 untill 2012 photovoltaic resin dye layer type electrode sheet sensitized oxide electrode contact polyester dye sensitized silicon silicon tab sealing sensitized solar subcell semiconductor substrate layer protection sheetdye sensitized sojulanrction etching module copolymer sensitized solar ctrealnl sparent condwucirtiinvge photovoltaic cellsolar cell moduldeye sensitized socloanr dcueclltive surface portion sheet for solar ceelllectrolyte transparent portion material sheet for solar electrode semiconductor substrate conductive cell module sensitizing solar subcell impurity C1 untill 2011 C2 untill 2011 C3 untill 2011 C4 untill 2011 C5 untill 2011 layer resin dye transparent organic photovoltaic sheet sensitized film dye electrode polyester dye sensitized oxide photoactive module solar cell modulesensitized solar layer electrode contact sealing dye sensitized soslialircon layer photovoltaic cellcell module sensitized solar ceelellctrode material forming sheet for solar cedlyle sensitized sotrlanr scpeallrent condeulcetcitvreon subcell sheet for solar electrode conductive comprises surface copolymer sensitizing amorphous photovoltaic tab module electrolyte plasma sensitized Until 2010 ? Until 2009 ? ? Until 2010 ? Until 2009 ? ? Until 2012 Until 2011 Until 2012 Until 2011
  • 13. Semantic Similarites between Clusters 0.6 0.5 0.4 0.3 0.2 0.1 0 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 Patent Academic Paper C1 C2 C3 C4 C5 Academic Paper until 2012 C1 C2 C3 C4 C5 Patent until 2012 Legend 0 0.6 Cosine Similarity
  • 14. Maximum Flow in Network 14 Node: Cluster Link: Semantic Similarity t t 0 t 1 t 1 +1 t end Meta Network Example) Source Sink
  • 15. Technology Layer (Patented Technologies) 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Science Layer (Academic Papers) Source Sink C5 C4 C3 C2 C1 C1 C2 C3 C4 C5
  • 16. Technology Layer (Patented Technologies) 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 C5 C1 C2 Source Sink C4 C3 C2 C1 C3 C4 C5 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Science Layer (Academic Papers)
  • 17. 17 Top 10 Keywords in each boundary cluster Science Layer Technology Layer Cluster 2. in Academic Paper 学until 術2005 論文(2005. C2) 特許公報(2007. C2) tio2 dye dye sensitized polymer dye sensitized sensitized dye sensitized solar dye sensitized dye sensitized solar cell sensitized solar sensitized solar cell sensitized solar cell sensitized solar nanocrystalline sensitizing dye sensitized solar pigment dye sensitized solar cell electrode Cluster 2. in Patent until 2007
  • 18. Summary ? It is necessary to understand knowledge relationships between Science and Technology, for Innovation Strategy. ? We considered the relationship as a time expanded and Heterogeneous network. ? We also considered this problem can be interpreted as a Maximum flow problem of Semantic Similarity Network. ? Photovoltaic field as case study. ? Grasped boundary clusters related to “Dye sensitized” field. 18
  • 19. Thank you for your attention. References ? Zucker, L.G., Darby, M.R.,“Capturing technological opportunities via Japan’s star scientists.”, Journal of Technology Transfer 26, 37–58, 2000. ? Hall B.H., Link A.N., Scott J.T., “Barriers inhibiting industry from partnering with universities: evidence from the advanced technology program.”, Journal of Technology Transfer 26, 87–98, 2001. ? Azoulay P., Ding W., Stuart T., 2006. “The impact of academic patenting on the rate, quality and direction of (public) research.”, NBER working paper 11917, Cambridge, MA., 2006. ? Rosenberg N., “Chemical engineering as a general purpose technology. In: Helpman, E. (Ed.), General Purpose Technologies and Economic Growth.”, MIT Press, Cambridge, pp. 167–192, 1998. ? Francis Narin, Kimberly S. Hamilton and Dominic Olivastro, “The increasing linkage between U.S. technology and public science”, Research Policy, Volume 26, Issue 3, October 1997, Pages 317–330, 1997. ? N. Shibata, Y. Kajikawa, and I. Sakata. “Extracting the commercialization gap between science and technology - case study of a solar cell.”, Technological Forecasting and Social Change, 77:1147–1155, 2010. ? Small H., “Citation Structure of an Emerging Research Area: Organic Thin Film.”, Proceedings of ISSI, pp. 718-725. 2007. ? 七丈, “共引用クラスタリングによる研究分野の動的把握に向けた試論”, 情報知識学会誌2013, Vol.23, No.3, 371-379, 2013. ? F.G. Engineer, G.L. Nemhauser, and M.W.P. Savelsgergh, “Dynamic programming-based column generation on Time-Expanded Network: Application to the Dial-a-Flight problem”, INFORMS Journal on Computing, 23(1), pp. 105-119, 2011. ? N. Shah, S. Kumar, F. Bastani, and I.L. Yen, “Optimization models for assessing the peak capacity utilization of intelligent transportation systems”, European Journal of Operational Research, 216, pp. 239-251, 2012. ? Newman M.E.J., “Fast algorithm for detecting community structure in networks”, Physical Review E, Vol. 69, p. 066133, 2004. ? Ino, H, Kudo. M, Nakamura. A, “A Comparative Study of Algorithms for Finding Web Communities”, Data Engineering Workshops, 2005. 21st International Conference, 1257, 2005. ? Horiike. T, Takahashi. Y, Kuboyama. T and Sakamoto. H, “Extracting Research Communities by Improved Maximum Flow Algorithm”, Knowledge-Based and Intelligent Information and Engineering Systems Lecture Notes in Computer Science Volume 5712, pp472-479, 2009. ? Victor Str?ele, Geraldo Zimbr?o, Jano M. Souza, “Modeling, Mining and Analysis of Multi-Relational Scientific Social Network”, Journal of Universal Computer Science, vol. 18, no. 8 (2012), 1048-1068, 2012. ? A. V. Goldberg and R. E. Tarjan, “A New Approach to the Maximum Flow Problem”, Journal of the ACM 35:921-940, 1988. ? Péter ?rdi, Kinga Makovi, Zoltán Somogyvári, Katherine Strandburg, Jan Tobochnik, Péter Volf, László Zalányi, “Prediction of Emerging Technologies Based on Analysis of the U.S. Patent Citation Network”, Scientometrics: Volume 95, Issue 1 (2013), Page 225-242, 2013.

Editor's Notes

  1. We can find lot of resarrch saing that To close relationship with science and technology will have posityve efffect. According to Dr. Narin, the rate of the number of patent which cite academic papers had rapidly grown from 1987 to 1994. It means science knowledge are effect to technology field is growing more and more.
  2. In the filed of DNA research, when Dr. Watthon and DrClis discovered 二重螺旋構造, the academic paper related to dna had published around 100 papers in a year. May be you can read all of paper with in the year. Now we can find annually 10万本(100thouand) paper related to the field from academic data base. Any resarecher cannot g read all of the paper. Knowledge is rapidly growing and more segmented. It si difficult ot gras all figure in each field even spacialist.
  3. Citaion network anaysls is used as one of the methodology in bebliometric analayasi. Network is consited with Node and Link. In this analysis we consider acadmiec paper and patent as NODE and consicer Citation relationship as Link.
  4. We deveided 2 process one is science layer, the other on is tehcnolgy layer. 1st of all we extracted dataset in each layere. In science layer, we extarced adacemic paper related to targeg filed). In tehnoogy layer, we extraced Pateng In both of layers, we proceed common 6 process.
  5. Sustainable and renewable energies have been widely accepted as a key concept for our common future [25]. A solar cell or photovoltaic cell, which is a device that converts solar energy into electricity via the photovoltaic effect, represents a promising research front for our future sustainable ecosystem.
  6. Newman's algorithm extracts tightly knit clusters with a high density of links within the cluster. The clustering algorithmis based on the idea of the maximization of modularity. Modularity is defined as the fraction of links that fall within clusters, minus the expected value of the same quantity if the links fall at random without regard for the clustered structure of that network. A high value of modularity represents a good division of clusters where only dense clusters remained within clusters and sparse links between clusters. Newman's algorithm [23] can efficiently find the point to maximize modularity over all possible divisions by cutting off links which connect clusters sparsely and extract clusters within which nodes are connected densely.
  7. tfi,d is the number of occurrences of ith term in document d, dfi is the number of documents containing ith term, N is the total number of documents
  8. This visualization shows the rsult of citation netrok in academic resarech related to Solar cell publicedh untill 2012
  9. We utlize cosine similairty to mearsures semantic similarity between clusters. This heatmap syos the cosicne similarity between the paper untill 2012 and patent antill 2012 You can find most high similar pair are C3 in paper and C3 in patent C3 is the cluster mainly focusec on It was also pointed out that research on dye-sensitized solar cells is focusing on the improvement in cell performance to a conversion efficiency of 15% through the development of new dyes and advanced cell structures, as well as that of production technology for large-area modules with integrated circuits on various substrates We donsidered C3 in Patent As Growing fields. Our problem set to dectech nowoledge transfere to Cluster 3 in patent.
  10. We concencd this problem as masimum flow ploblemn in network. That is meta network of citation network. What we analyze in Maximum flow is how many flow units each node can pass to each other. Based on Figure 3, we can see that node 1 can pass up to ten units to nodes 2, 3, and 4. Node 2 can pass up to 5 units to both nodes 6 and 7. Node 3 can pass up to six flow units to node 6, and four units to node 7. Node 4 can pass up to ten flow units to node 5. Finally, nodes 5, 6 and 7 can pass up to ten, eleven and nine units to receptor node 8, respectively. Thus, the maximum flow from node 1 to node 8 is of 30 units.” “The objective of the developed algorithm is to group nodes with the largest flow of knowledge between them.” In this network, we cnceden Cluster in each year as and Similarity as Link Victor Str啼le, Geraldo Zimbr黍, Jano M. Souza 2011, “Evaluating Knowledge Flow in Multirelational Scientific Social Networks”
  11. We assume 2 heterogenous network.
  12. We regared these 2cluster is boundary clusters dye sensitized knolwege had been transit from scienct to technology.
  13. Zucker, L.G., Darby, M.R.,“Capturing technological opportunities via Japan’s　star scientists.”, Journal of Technology Transfer 26, 37–58, 2000. Hall B.H., Link A.N., Scott J.T., “Barriers inhibiting industry from partnering　with universities: evidence from the advanced technology program.”, Journal of　Technology Transfer 26, 87–98, 2001. Azoulay P., Ding W., Stuart T., 2006. “The impact of academic patenting on the　rate, quality and direction of (public) research.”, NBER working paper 11917, Cambridge,　MA., 2006. Rosenberg N., “Chemical engineering as a general purpose technology. In: Helpman,　E. (Ed.), General Purpose Technologies and Economic Growth.”, MIT Press,　Cambridge, pp. 167–192, 1998. Francis Narin, Kimberly S. Hamilton and Dominic Olivastro, “The increasing linkage between U.S. technology and public science”, Research Policy, Volume 26, Issue 3, October 1997, Pages 317–330, 1997. N. Shibata, Y. Kajikawa, and I. Sakata. “Extracting the commercialization gap between science and technology - case study of a solar cell.”, Technological Forecasting and Social Change, 77:1147–1155, 2010. Small H., “Citation Structure of an Emerging Research Area: Organic Thin Film.”, Proceedings of ISSI, pp. 718-725. 2007. 七丈, “共引用クラスタリングによる研究分野の動的把握に向けた試論”, 情報知識学会誌 2013, Vol.23, No.3, 371-379, 2013. F.G. Engineer, G.L. Nemhauser, and M.W.P. Savelsgergh, “Dynamic programming-based column generation on Time-Expanded Network: Application to the Dial-a-Flight problem”, INFORMS Journal on Computing, 23(1), pp. 105-119, 2011. ? N. Shah, S. Kumar, F. Bastani, and I.L. Yen, “Optimization models for assessing the peak capacity utilization of intelligent transportation systems”, European Journal of Operational Research, 216, pp. 239-251, 2012. Newman M.E.J., “Fast algorithm for detecting community structure in networks”, Physical Review E, Vol. 69, p. 066133, 2004. Ino, H, Kudo. M, Nakamura. A, “A Comparative Study of Algorithms for Finding Web Communities”, Data Engineering Workshops, 2005. 21st International Conference, 1257, 2005.? Horiike. T, Takahashi. Y, Kuboyama. T and Sakamoto. H, “Extracting Research Communities by Improved Maximum Flow Algorithm”, Knowledge-Based and Intelligent Information and Engineering Systems Lecture Notes in Computer Science Volume 5712, pp472-479, 2009. Victor Str?ele, Geraldo Zimbr?o, Jano M. Souza, “Modeling, Mining and Analysis of Multi-Relational Scientific Social Network”, Journal of Universal Computer Science, vol. 18, no. 8 (2012), 1048-1068, 2012. A. V. Goldberg and R. E. Tarjan, “A New Approach to the Maximum Flow Problem”, Journal of the ACM 35:921-940, 1988. Péter ?rdi, Kinga Makovi, Zoltán Somogyvári, Katherine Strandburg, Jan Tobochnik, Péter Volf, László Zalányi, “Prediction of Emerging Technologies Based on Analysis of the U.S. Patent Citation Network”, Scientometrics: Volume 95, Issue 1 (2013), Page 225-242, 2013.