A Method Of Technology Opportunity Discovery Based On Knowledge Network

Discovering and grasping technology opportunities is of great significance to the establishment of an innovative country.The timely discovery of technology opportunities that are at the forefront of technology and feasible is the foundation of enterprise innovation,and often plays a vital role in the survival and development of enterprises.In the era of big data,technical information is hidden in the ocean of data and knowledge.It is more difficult to accurately grasp the status quo of technology and discover potential and valuable technology opportunities.Moreover,there are difficulties in selecting and evaluating many potential technology opportunities.There are three main problems in these technology opportunity discovery methods: first,the representation form of technology opportunity is usually technology domain or sub-domain(with larger granularity),while finding more specific and easy-to-interpret technology opportunity usually requires more intervention of domain experts;second,most of the existing researches use a few opportunity characteristics(such as "blank","outlier" or "similarity")to identify technology opportunities,and may abandon some technology opportunities that do not conform to these characteristics but are also valuable;third,although existing research has put forward many criteria for evaluating technology opportunities,most of these criteria lack empirical support.In order to solve the above problems,a technology opportunity discovery method based on knowledge network construction,evaluation and search is proposed.Firstly,applying the knowledge network theory,the domain knowledge network is constructed by taking the technical words in the title and abstract of patents in the domain as knowledge elements(i.e.nodes in the knowledge network)and the semantic dependency between words as edges,and its connected sub-networks are regarded as candidate technology opportunities.This domain knowledge network preserves the importance and novelty information of knowledge elements(nodes)and their relationships(edges)in the domain.It can also save the information of node and edge source patents as attributes to help analyze and evaluate technology opportunities.Secondly,the opportunity value evaluation is carried out according to the risk assessment theory and method,the decision variables that have influence on the opportunity value are designed with reference to the existing research,and the effect of these decision variables on the opportunity value is empirically tests according to historical technology opportunities.Then,the decision variables that have influence on the value of technology opportunities are further expanded into characteristic indicators for searching opportunities,constructing the objective function of opportunity searching,and the heuristic optimization algorithm is used to carry out technology opportunities in the overall domain knowledge network to obtain the optimal set of technology opportunities.Finally,the method is applied to the technical field of "measuring mechanical vibration or sound waves by electromagnetic means",and the feasibility and the effectiveness of the method for opportunity discovery are empirically tested.The technology opportunity discovery method proposed in this paper provides a new solution to the above three problems in the existing methods.First,by constructing a domain knowledge network with complete technical information and expressing technology opportunities in the form of knowledge subnetworks,the expression of opportunities is more detailed and specific,easy to interpret,and reduces the degree of expert intervention to solve the first problem;then,using a process different from the previous technology opportunity discovery and evaluation,that is,the historical technology opportunity is used to empirically test the opportunity characteristics that affect the value of the opportunity,and the evaluation indicators are obtained and used to search for future technology opportunities,so that the evaluation of technology opportunities is more objective and empirical,and provide new ideas for the evaluation of technology opportunities,so as to solve the third problem;finally,applying the heuristic optimization algorithm to search for technology opportunities in the whole field,which avoids the limitation that a few opportunity features exclude some valuable opportunities,and increases the depth and breadth of opportunity discovery,thereby solving the second problem.The research results will provide more sophisticated technology opportunity discovery and evaluation tools for the company's R & D activities.