Early detection of technology opportunity based on analogy design and phrase semantic representation

In order to gain competitive advantage, technology opportunity detection in the latest and fast-growing areas has been becoming an important research issue. However, current research on technology opportunity detection is often focus on verifying the technology opportunities that have occurred, using the accumulated data from a specific field. Because of the time needed for data accumulation, these methods often have a substantial time lag and hard to early detect technology opportunities. It also leads to challenges to explore technology opportunities which still have not been covered in the current dataset. Moreover, phrase has more semantics than words but still rarely used and semantic represented in the process of technology opportunity detection. Therefore, this paper proposes a method based on analogy design and phrase semantic representation for early detection of technology opportunity. Firstly, the source field corresponding to target field for analogy design is carefully selected, thus indirectly expanding the data coverage of the target field through the data from the source field. Secondly, effect phrases in both source field and target field are automatically extracted by BiLSTM-CRF and semantic represented by representation learning, then the analogy relationships are established through topic clustering on overall data. Finally, the scores of the topics are calculated based on ODI (outcome-driven innovation) and the topics with a high score are considered as early detected technology opportunities. The proposed method is validated using analogy between 3G and 4G. In this process, 3G is used as the source and 4G patents published in the early stage are used as the target for detecting technology opportunities in 4G, and the rest 4G patents published in the later stage are used for detecting the actual evolution results of technology opportunities. The comparison results prove that every detected technology opportunity in the early stage matches one or more topics of the actual evolution results in the later stage. In addition, this paper uses analogy between 4G (source field) and 5G (target field) for technology opportunities prediction, which may provide useful and helpful results for decision making in 5G and a good example for further application in other areas. These results have proved that the proposed method is effective and useful. Simultaneously, this method is a preliminary research and still need to be further studied on other datasets with different analogy types.

[1]  Eduard H. Hovy,et al.  End-to-end Sequence Labeling via Bi-directional LSTM-CNNs-CRF , 2016, ACL.

[2]  Michael A. Hitt,et al.  Institutional Ownership Differences and International Diversification: The Effects of Boards of Directors and Technological Opportunity , 2003 .

[3]  Ashok K. Goel,et al.  Use of design patterns in analogy-based design , 2004, Adv. Eng. Informatics.

[4]  Kevin Otto,et al.  Design-by-analogy: experimental evaluation of a functional analogy search methodology for concept generation improvement , 2015 .

[5]  Jonghwa Kim,et al.  Technology opportunity discovery (TOD) from existing technologies and products: A function-based TOD framework , 2015 .

[6]  Julie S. Linsey.  Design-by-analogy and representation in innovative engineering concept generation , 2007 .

[7]  Byungun Yoon,et al.  A systematic approach for identifying technology opportunities: Keyword-based morphology analysis , 2005 .

[8]  Chanwoo Cho,et al.  An Empirical Analysis on Purposes, Drivers and Activities of Technology Opportunity Discovery: The Case of Korean SMEs in the Manufacturing Sector , 2016 .

[9]  Oh-Jin Kwon,et al.  Product opportunity identification based on internal capabilities using text mining and association rule mining , 2016 .

[10]  Guijun Wang,et al.  A Study of Chinese Document Representation and Classification with Word2vec , 2016, 2016 9th International Symposium on Computational Intelligence and Design (ISCID).

[11]  Li Jiang,et al.  Empirical research on the technology opportunities analysis based on morphology analysis and conjoint analysis , 2008, PICMET '08 - 2008 Portland International Conference on Management of Engineering & Technology.

[12]  Duk Hee Lee,et al.  Patent citation network analysis for the domain of organic photovoltaic cells: Country, institution, and technology field , 2013 .

[13]  Sungjoo Lee,et al.  An approach to discovering new technology opportunities: Keyword-based patent map approach , 2009 .

[14]  M. Nieto,et al.  Absorptive capacity, technological opportunity, knowledge spillovers, and innovative effort , 2005 .

[15]  Jeffrey S. Pinegar.  What Customers Want: Using Outcome‐Driven Innovation to Create Breakthrough Products and Services by Anthony W. Ulwick , 2006 .

[16]  Thomas Demeester,et al.  Representation learning for very short texts using weighted word embedding aggregation , 2016, Pattern Recognit. Lett..

[17]  Kwangsoo Kim,et al.  Detecting signals of new technological opportunities using semantic patent analysis and outlier detection , 2011, Scientometrics.

[18]  Myong Kee Jeong,et al.  Two-phase edge outlier detection method for technology opportunity discovery , 2017, Scientometrics.

[19]  Li Xin,et al.  Detecting Technology Opportunities Based on Papers and Patents for Perovskite Solar Cells , 2016 .

[20]  Bokyoung Kang,et al.  Novelty-focused patent mapping for technology opportunity analysis , 2015 .

[21]  Uwe Aickelin,et al.  CRNN: A Joint Neural Network for Redundancy Detection , 2017, 2017 IEEE International Conference on Smart Computing (SMARTCOMP).

[22]  Dafna Shahaf,et al.  Accelerating Innovation Through Analogy Mining , 2017, KDD.

[23]  Shih-Chieh Fang,et al.  Exploring technological opportunities by mining the gaps between science and technology: Microalgal biofuels , 2015 .

[24]  Yuantao Gu,et al.  Cross-Label Suppression: A Discriminative and Fast Dictionary Learning With Group Regularization , 2017, IEEE Transactions on Image Processing.

[25]  Christopher L. Magee,et al.  Exploring technology opportunities by visualizing patent information based on generative topographic mapping and link prediction , 2018, Technological Forecasting and Social Change.

[26]  Jeffrey P. Bigham,et al.  Combining Independent Modules to Solve Multiple-choice Synonym and Analogy Problems , 2003, ArXiv.

[27]  Michael A. Orloff,et al.  Inventive Thinking through TRIZ , 2003 .

[28]  Zhiyuan Liu,et al.  Joint Learning of Character and Word Embeddings , 2015, IJCAI.

[29]  Julie Linsey,et al.  Transforming functional models to critical chain models via expert knowledge and automatic parsing rules for design analogy identification , 2017, Artificial Intelligence for Engineering Design, Analysis and Manufacturing.

[30]  W. Schnotz,et al.  Surface and deep structures in graphics comprehension , 2014, Memory & Cognition.

[31]  Alan L. Porter,et al.  Technology opportunities analysis , 1995 .

[32]  Kwangsoo Kim,et al.  Creating patents on the new technology using analogy-based patent mining , 2014, Expert Syst. Appl..

[33]  Alexander Panchenko,et al.  How much does a word weigh? Weighting word embeddings for word sense induction , 2018, ArXiv.

[34]  Jaewoong Choi,et al.  Technology opportunity discovery under the dynamic change of focus technology fields: Application of sequential pattern mining to patent classifications , 2019, Technological Forecasting and Social Change.

[35]  Jonathan Cagan,et al.  ON THE EFFECTIVE USE OF DESIGN-BY-ANALOGY: THE INFLUENCES OF ANALOGICAL DISTANCE AND COMMONNESS OF ANALOGOUS DESIGNS ON IDEATION PERFORMANCE , 2011 .

[36]  Julie Linsey,et al.  Establishing functional concepts vital for design by analogy , 2015, 2015 IEEE Frontiers in Education Conference (FIE).

[37]  Jeongjin Lee,et al.  Technology opportunity discovery to R&D planning: Key technological performance analysis☆ , 2017 .

[38]  Anthony W. Ulwick.  What Customers Want: Using Outcome-Driven Innovation to Create Breakthrough Products and Services , 2005 .

[39]  Richard Catrambone,et al.  The effects of surface and structural feature matches on the access of story analogs. , 2002, Journal of experimental psychology. Learning, memory, and cognition.

[40]  Robert L. Goldstone,et al.  Relational similarity and the nonindependence of features in similarity judgments , 1991, Cognitive Psychology.

[41]  J. Paris,et al.  The Counterpart Principle of Analogical Support by Structural Similarity , 2014 .

[42]  Ola Olsson,et al.  Technological Opportunity and Growth , 2005 .

[43]  S. Kanmani,et al.  Document clustering and topic discovery based on semantic similarity in scientific literature , 2011, 2011 IEEE 3rd International Conference on Communication Software and Networks.

[44]  Paul Geroski,et al.  INNOVATION, TECHNOLOGICAL OPPORTUNITY, AND MARKET STRUCTURE , 1990 .

[45]  Jeffrey Dean,et al.  Distributed Representations of Words and Phrases and their Compositionality , 2013, NIPS.

[46]  Sungjoo Lee,et al.  Discovering new technology opportunities based on patents: Text-mining and F-term analysis , 2017 .

[47]  Quoc V. Le,et al.  Distributed Representations of Sentences and Documents , 2014, ICML.

[48]  R. Tan,et al.  Design by Analogy: Achieving More Patentable Ideas from One Creative Design , 2018, Chinese Journal of Mechanical Engineering.

[49]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[50]  Jun Guo,et al.  A novel negative sampling based on TFIDF for learning word representation , 2016, Neurocomputing.

[51]  Bo T. Christensen,et al.  The relationship of analogical distance to analogical function and preinventive structure: the case of engineering design , 2007, Memory & cognition.