Mining Data from the Congressional Record

Zhengyu Ma,Tianjiao Qi,James Route,Amir Ziai
DOI: https://doi.org/10.48550/arXiv.1906.00529
2019-06-03
Abstract:We propose a data storage and analysis method for using the US Congressional record as a policy analysis tool. We use Amazon Web Services and the Solr search engine to store and process Congressional record data from 1789 to the present, and then query Solr to find how frequently language related to tax increases and decreases appears. This frequency data is compared to six economic indicators. Our preliminary results indicate potential relationships between incidence of tax discussion and multiple indicators. We present our data storage and analysis procedures, as well as results from comparisons to all six indicators.
Information Retrieval,Computers and Society
What problem does this paper attempt to address?