Research of Ambiguity checking by NLP in EIA Report

Abstract

The environment impact assessment(EIA) report is an important reference measure for measuring the overall environmental status of the region. Both the data and the text description in the report are intuitive manifestations of the environmental status. Therefore, the consistency of the text description and data is very important in the EIA report. This paper focus on text description and data ambiguity detection of atmospheric indices by using natural language processing(NLP) techniques in EIA report. The key issue of detection is matching. In the process, firstly taking the EIA of a project in Haidian District of Beijing as an example source data and establishing a corpus of atmospheric EIA. The corpus contains atmospheric environmental indicators, the degree of description of these indices, values, etc. Secondly, comparing the subjective and objective weighting method and choose a comprehensive weight method to match target for ambiguity detection. Finally, using regular expression to match the text description and data of atmospheric indices in EIA report. This method of matching can reduce the complexity compared with traditional matching method.