This study proposes an integrated trend analysis methodology by applying bibliometric analysis and text mining on both scientific paper and patent database. The bibliometric analysis is investigated in four different dimensions: publication growth, country/region, organization, classification. However, the text mining is obtained by keyword analysis for different time periods and research field maps. The integrated method is applied on etching technology which is a material processing technology particularly important for semiconductor industry. On the basis of the results obtained in the practice of etching technology, six different relations among the proposed two publication database (paper and patent) and two conventional trend forecasting method (bibliometrics and text-mining) are systematically analyzed as a framework for depicting a desirable Sci-Tech trend analysis model: Bibliometric analysis should be investigated first, and then text-mining is subsequently applied to discover important research topics of selected fields. The importance of paper and patent should be equally treated for obtaining a complete result taking into both theory and practice into consideration.