You can view the full text of this article for free using the link below.

Title: An ex-convict recognition method based on text mining

Authors: Mingyue Qiu; Xueying Zhang; Xinmeng Wang

Addresses: School of Information Technology, Nanjing Forest Police College, Nanjing, 210023, China ' Key Laboratory of Virtual Geographic Environment, Nanjing Normal University, Nanjing, 210023, China ' School of Information Technology, Nanjing Forest Police College, Nanjing, 210023, China

Abstract: Currently, a large proportion of existing cases in the grassroots public security organisations were committed by ex-convicts. Grassroots police officers cannot directly and rapidly judge whether a suspect is an ex-convict who has committed a case. To solve this problem, an attempt is made to analyse the case report data in a branch bureau in 2021 through data mining. Using the brief case texts in the case report data as the data source, different models based on various algorithms were established to judge whether the ex-convict committed the case. Next, using different algorithms, the ex-convict in the database was ascertained based on the similarity degree results. Finally, the similarity results (the highest similarity reached 94.8) using different methods were calculated, added, and ranked in descending order to submit an ex-convict list to the grassroots police officers for further artificial judgement. Accordingly, grassroots police officers can conduct rapid recognition of ex-convicts when a case is reported. The present model is tested well in the actual applications in the local police stations, suggesting that the model can provide overwhelming support in the daily work of police stations, and with the mutual cooperation and gradual promotion among the police stations, large amounts of human and material resources can finally be saved.

Keywords: natural language processing; text mining; similarity analysis; recognition of people with previous conviction.

DOI: 10.1504/IJSN.2023.129905

International Journal of Security and Networks, 2023 Vol.18 No.1, pp.10 - 18

Received: 21 Jun 2022
Accepted: 27 Jun 2022

Published online: 03 Apr 2023 *

Full-text access for editors Full-text access for subscribers Free access Comment on this article