Abstract:

We gave a general feature set of spatial relations based on qualitative spatial reasoning of spatial relations, and combining with network crawling, text mining, spatial database and machine learning methods, we proposed a general recognizing framework to recognize most of the spatial relations in Chinese text, and solved the problem of spatial relation recognition in Chinese text of big data. Compared with traditional spatial relation models, the proposed method is suitable for dealing with spatial information in Chinese text of big data, and obtains a higher recognition rate.

Key words: qualitative spatial reasoning, spatial relation, general feature, machine learning