Data Scraping to Leverage AI Precision in Pure Science

Abstract
Applied into Physics, Chemistry, and other scientific fields, Natural Language Process Model plays an important role in bridging Logical Reasoning with Optical Character Recognition, data retrieval, and outsourced calculator. In the real world, the uncertainty or less than 100% accuracy of AI limits its widespread applications in academia at the current stage. However, embedded with consulting techniques, data precision will be greatly improved, where some key factors can be manually put into equations or formulas, such as physical constants, unit conversion, and so forth, even as precise as to the 7th decimal place of pi. While an automatic search in one sentence can reduce spatial stories into tabular data, it is inevitable that a particular webpage is either not found or mixed with words and pictures, but rather nothing can deal with sentence parsing or word separating, completely clean. By introducing a hybrid of automation and manual work, it takes only a few minutes for data to be filtered, extracted, and predicted, a workflow that could have cost a long time otherwise. Overall, this paper is aimed at a precise calculation for pure science and for the sake of security in practical uses across various industries.