Natural language processing-based detection of systematic anomalies among the narratives of consumer complaints
Peiheng Gao et al.
What the paper says
We develop a natural language processing-based procedure for detecting systematic nonmeritorious consumer complaints – called simply systematic anomalies – among complaint narratives. While classification algorithms are used to identify meritorious complaints, these algorithms may falter in the case of smaller and more frequent systematic patterns of nonmeritorious complaints. This could be for a variety of reasons, such as technical issues or the natural limitations of human analysts. Therefore, following the classification stage, the complaint narratives are converted into quantitative data, which are then analyzed using indexes for detecting systematic anomalies. An illustration of the entire procedure is provided using complaint narratives from the Consumer Complaint Database of the US Consumer Financial Protection Bureau. The results suggest that the support vector machine outperforms other selected classifiers. Although the classification results with the Valence Aware Dictionary for sEntiment Reasoning (VADER) intensity pertinent to the featurization step have lower accuracy, they contain fewer nonmeritorious complaints than those without the VADER intensity.
2 citations
Evidence weight
Balanced mode · F 0.40 / M 0.15 / V 0.05 / R 0.40
| F · citation impact | 0.22 × 0.4 = 0.09 |
| M · momentum | 0.55 × 0.15 = 0.08 |
| V · venue signal | 0.50 × 0.05 = 0.03 |
| R · text relevance † | 0.50 × 0.4 = 0.20 |
† Text relevance is estimated at 0.50 on the detail page — for your query’s actual relevance score, open this paper from a search result.