Textual Knowledge Visualization with Artificial Intelligence Optimization Approach, FAQ as an Example

• Main Authors: HUANG, CHIN-FU, 黃進福
• Other Authors: CHIU, DENG-YIV
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/bx4k43

碩士 === 中華大學 === 資訊管理學系 === 105 === For organizations or companies, the amount of information created day by day is very amazing. However, no matter the internal operational transfer information or external operational transfer information, these information is the knowledge assets for organizations or companies. The company's customer service is very important. The problem records can help company to understand customer's impressions. Company can use the problem records to develop solutions to improve the shortcomings. But with the large problem records it is not easy to find the above information. Therefore, this study use the information retrieval's text mining to establish topic knowledge map for a company's customer problem records and explore the hidden information by topic knowledge maps. The establishment of topic knowledge map is mainly to transfer topic knowledge structure into two-dimensional space. Topic knowledge structure is through text mining of information retrieval to explore a large number of documents and then find important keywords to be the topics. The words highly correlated with topics will be find too. Then topic knowledge structure will be established by these topics and words. But the topic knowledge structure mainly is used to explore hierarchy of knowledge, so it is difficult to find correlation or influence between knowledge. Therefore, it is necessary to visualize the knowledge structure into the knowledge map to explore correlation or influence between knowledge. In order to establish the topic knowledge map, we use transformation coordinate matrix of multidimensional scale method to keep relation between objects. Therefore, the design of transformation coordinate matrix is very important to keep relation between objects before and after transfer. However, this study will try to obtain the optimal transformation coordinate matrix with genetic algorithm. Finally, this study will try to understand the relation and influence between topics with the topic knowledge map of customer problem records.