Management of Unstructured Geological Data Based on Hadoop

For over a century Geological Survey of China (CGS) has accumulated an enormous amount of data on geology, mineral resources and petroleum fields of China. Many of this geological information were recorded in unstructured formats. How to derive useful knowledge from these data for industrial and research users is a challenging problem for a long period. In this paper, we developed a hierarchical model to store the unstructured geological data based on Hadoop. Using a modified KNN (K-Nearest Neighbor) algorithm based on MapReduce framework, top 10 geological words were identified and analyzed in the unstructured geological test data.