양방향 LSTM기반 시계열 특허 동향 예측 연구
A patent application filing forecasting method based on the bidirectional LSTM
1
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
특정 분야의 특허출원수는 기술의 수명주기 및 산업의 활성화 정도와 밀접한 관계를 가지고 있다. 따라서 사전에 사업을 준비하는기업들과 미래 유망 기술을 초기 단계에서 선발하여 투자하고자 하는 정부 기관들은 미래의 특허 출원수 예측에 대해 큰 관심을 가지고 있다. 본 논문에서는 시계열 데이터에 적합한 RNN의 기법 중 하나인 양방향 LSTM 기법을 이용하여 기존 예측 방법들보다정확도를 높이는 방법을 제안한다. 5개 분야의 대한민국 특허 출원 데이터에 대해서 제안된 방법은 기존에 사용되던 확산 모델 중하나인 Bass 모델과 비교하여 평균 절대 백분율 오차(MAPE)의 값이 약 16퍼센트 향상된 결과를 보여준다.
Abstract
The number of patent application filing for a specific technology has a good relation with the technology’s lifecycle and future industry development on that area. So industry and governments are highly interested inforecasting the number of patent application filing in order to take appropriate preparations in advance. In thispaper, a new method based on the bidirectional long short-term memory(LSTM), a kind of recurrent neuralnetwork(RNN), is proposed to improve the forecasting accuracy compared to related methods. Compared with theBass model which is one of conventional diffusion modeling methods, the proposed method shows the 16% higherperformance with the Korean patent filing data on the five selected technology areas.
참고문헌 (19)
- R. Haupt, "Patent indicators for the technology life cycle development", Research Policy, 36(3), pp.387~398 (2007)[DOI]
- Chen-Yuan Liu, "Forecasting the development of the biped robot walking technique in Japan through S-curve model analysis", Scientometrics, 82(1), pp.21~36 (2009), Springer Science and Business Media LLC[DOI]
- Peter Hingley, "Methods for forecasting numbers of patent applications at the European Patent Office", World Patent Information, 26(3), pp.191~204 (2004), Elsevier BV[DOI]
- 김도회, "특허정보를 활용한 기술 확산 예측: NCW 정보보호기술을 중심으로", 한국콘텐츠학회 논문지, 9(2), pp.125~132 (2009), 한국콘텐츠학회
- 홍정식, "NLS와 OLS의 하이브리드 방법에 의한 Bass 확산모형의 모수추정", 대한산업공학회지, 37(1), pp.74~82 (2011), 대한산업공학회
- 김갑조, "특허 토픽 모델링과 성장주기곡선을 통한 유망기술 발굴", 한국지능시스템학회 논문지, 27(4), pp.357~363 (2017), 한국지능시스템학회
- 남대경, "토픽모델 및 특허분석을 통한 차량용 반도체 기술 추세 분석", 기술혁신학회지, 21(3), pp.1155~1178 (2018), 한국기술혁신학회
- Moses Ntanda Kyebambe, "Forecasting emerging technologies: A supervised learning approach through patent analysis", Technological Forecasting and Social Change, 125, pp.236~244 (2017), Elsevier BV[DOI]
- Raman Dutt, "Novel mixed-encoding for forecasting patent grant duration", World Patent Information, 64, pp.102007~ (2021), Elsevier BV[DOI]
- Gabjo Kim, "A novel approach to forecast promising technology through patent analysis", Technological Forecasting and Social Change, 117, pp.228~237 (2017), Elsevier BV[DOI]
- Joon Hyung Cho, "Predicting future technological convergence patterns based on machine learning using link prediction", Scientometrics, 126(7), pp.5413~5429 (2021), Springer Science and Business Media LLC[DOI]
- C. Lee Giles, "Dynamic recurrent neural networks: Theory and applications", IEEE Transactions on Neural Networks, 5(2), pp.153~156 (1994), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Sepp Hochreiter, "Long Short-Term Memory", Neural Computation, 9(8), pp.1735~1780 (1997), MIT Press - Journals[DOI]
- Y. Bengio, "Learning long-term dependencies with gradient descent is difficult", IEEE Transactions on Neural Networks, 5(2), pp.157~166 (1994), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Alex Sherstinsky, "Fundamentals of Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) network", Physica D: Nonlinear Phenomena, 404, pp.132306~ (2020), Elsevier BV[DOI]
- M. Schuster, "Bidirectional recurrent neural networks", IEEE Transactions on Signal Processing, 45(11), pp.2673~2681 (1997), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Long short-term memory recurrent neural network architectures for large scale acoustic modeling, "H. Sak", International Speech Communication Association (INTERSPEECH), 338~342 (2014)
- "Understanding LSTM Networks"[URL]
- D. Lee, "Exploratory research on the analysis of national R&D programs using growth model", Korea Institute of Science and Technology Evaluation and Planning, 27 (2014)[DOI]