Yu-Ting Chiu

Visualization of Financial Trends Using Chance Discovery Methods

Due to the concern of business performance and the consideration of possible investment, stakeholders inside and outside the business need to understand the financial status of a company. Usually, reading and interpreting the public offering financial statements is a common way to obtain an overall financial situation of a company. Apart from reading them directly, some visualization methods were employed so as to assist the less-professional stakeholders to understand the financial status more intuitively and painlessly. Chance discovery is one of the visualization methods which may be suitable to be applied in the financial trend detection area. In this study, a financial data visualization framework has been proposed to produce the serial KeyGraph for explaining the financial situations of two companies. Subsequently, an integrated map of financial trend was generated for depicting the overview of the accumulative five-year scenario. Finally, the visualization of financial trend was attained using serial KeyGraph and integrated map.

Applying Chance Discovery with Dummy Event in Technology Monitoring of Solar Cell

One of the green energy, solar cell, is growing rapidly; the monitoring of its technological situation becomes critical for the stakeholders nowadays. Meanwhile, the patent data contains plentiful technological information from which is worthwhile for exploring further knowledge. Therefore, a graph-based approach, chance discovery, is employed so as to analyze the patent data, to form the technological scenarios, and to explain the overview of solar cell technology. Finally, the relationships between technology and companies, between application and companies have been observed, and the strategic suggestions have been proposed accordingly.

Using IPC-Based Clustering and Link Analysis to Observe the Technological Directions

To explore the technological directions of an industry is essential for companies and stakeholders to anticipate the future situations and R&D activities. Patent data contains plentiful technical information, which is appropriate to be used in technological analysis in order to find out the technical topics and possible directions. Due to the complex nature of patent data, two data mining methods: IPC-based clustering and link analysis, are used to figure out the potential tendencies on thin-film solar cell. An IPC-based clustering algorithm will be proposed and utilized to generate the significant categories via the IPC and Abstract fields, while the link analysis will be adopted to draw a link diagram for the whole dataset via the Abstract, Issue Date, and Assignee Country fields. During experiment, the technical categories will be identified using the IPC-based clustering, and the technological directions will be found through the link analysis. Finally, the recognized technical categories and technological directions will be provided to the managers and stakeholders for assisting their decision making.

To propose strategic suggestions for companies via IPC classification and association analysis

Strategic suggestions are essential for companies to facilitate the top management to foresee and review the future directions of their companys research and development investment. Therefore, a research design has been formed for performing the strategic planning on technology where IPC classification was employed to divide the patents into different categories and association analysis was adopted to discover the relations between terms and between clusters. Consequently, the visualized results, crystallized diagrams and integrated map, were generated and the relations between technical topics and companies were observed. Finally, according to the relations, the strategic suggestions on thin-film solar cell for companies were recognized and proposed.

Trend detection on thin-film solar cell technology using cluster analysis and modified data crystallization

Thin-film solar cell, one of green energies, is growing at a fast pace with its long-lasting and non-polluting natures. To detect the potential trends of this technology is essential for companies and relevant industries so that the competitive advantages of companies can be retained and the developing directions of industries can be perceived. Therefore, a research framework for trend detection has been formed where cluster analysis is employed to perform the similarity measurement, and data crystallization is adopted to conduct the association analysis. Consequently, the relation patterns were identified from the relations among companies, issue years, and techniques. Finally, according to the relation patterns, the potential trends of thin-film solar cell were detected for companies and industries.

An Experiment Model of Grounded Theory and Chance Discovery for Scenario Exploration

To explore the scenario of a technology is a valuable task for managers and stakeholders to grasp the overall situation of that technology. Solar cell, one of renewable energies, is growing at a fast pace with its unexhausted and non-polluted characters. In addition, the patent data contains plentiful technological information from which is worthwhile to extract further knowledge. Therefore, an experiment model has been proposed so as to analyze the patent data, to form the scenario, and to explore the tendency of solar cell technology. Finally, the relation patterns were identified, the directions of solar cell technology were recognized, the active companies and vital countries in solar cell industry were also observed.

Using text mining and chance discovery for exploring technological directions via patent data

To work out the situation and directions of a technology is essential for an industry, companies, and stakeholders so as to facilitate the decision makers to bring up rational judgments for their decisions. Solar cell, one of renewable energies, is growing at a fast pace with its long-lasting and non-polluting natures. In order to explore the situation and directions of technology, two approaches, namely text mining and chance discovery are employed to conduct the cluster analysis and association analysis on the patent data. Consequently, the technical topics have been found after similarity measurement; the subtopics and relations have been recognized after KeyGraph generation and data crystallization. Finally, according to the relation patterns, the situation and directions of thin-film solar cell have been identified and stated.

Chance discovery with data crystallization for scenario formation in solar cell technology

As the solar cell, one of renewable energies, is growing at a fast pace, the recognition of its technological situation becomes necessary for a company and stakeholders nowadays. Meanwhile, the patent data contains plentiful technological information from which is worth exploring to extract further knowledge. Therefore, a graph-based approach, chance discovery, is employed so as to analyze the patent data, to form the technological scenarios, and to explain the tendency of solar cell technology. Finally, several topics of solar cell technology have been identified, the directions of each topic have been depicted, and the relations between topics have been also observed.

Exploring Technology Opportunities in an Industry via Clustering Method and Association Analysis

To explore the technology opportunity is essential for a company and an industry so that the company can consider the allocation of R&D investments and the industry can observe the developing directions of rare topics. Patent data contains plentiful technological information from which it is worthwhile to extract further knowledge. Therefore, a research framework for exploring the technology opportunity has been formed where clustering method is employed to generate the clusters, similarity measurement is adopted to identify the variant patents, and association analysis is used to recognize the focused rare topics. Consequently, the clusters were generated and named, the variant patents were found, the focused rare topics were recognized, and the technology opportunities for companies were discussed. Finally, the variant patents and the technology opportunities would be provided to assist the decision makers of companies and industries.

Emerging technology exploration using rare information retrieval and link analysis

To explore the clues of an emerging technology is essential for a company or an industry so that the company can consider the feasibility of resource allocation to the technology and the industry can observe the developing directions of the technology. Patent data contains plentiful technological information from which it is worthwhile to extract further knowledge. Therefore, a research framework for emerging technology exploration has been formed where rare information retrieval is designed to sift out the rare patents, cluster analysis is employed to generate the clusters, and link analysis is adopted to measure the link strength between the rare patents and clusters. Consequently, the rare patents were found, the clusters were generated and named, the notable rare patents were recognized, and the potentiality of emerging technology was discussed. Finally, the notable rare patents and the potentiality of emerging technology would be provided to the decision makers of companies and industries.