執法語言智能訓練系統在反詐預警勸阻領域的應用研究
楊勇、陶魏光鍚、鄧媛

摘要：
電信網絡詐騙犯罪由於其作案手段的非接觸性和網絡犯罪的虛擬性，且缺少傳統意義上的物證載體，導致對其進行偵查打擊尤為困難。因此電信網絡詐騙的犯罪生態治理應當注重源頭治理，以犯罪預防為核心，從而在本源上降低發案率。電詐犯罪的預防分為預警和勸阻兩個部分，反詐預警以宣傳工作為主，而反詐勸阻主要是針對將要發生或是已經發生的詐騙案件進行人工干預。然而由於涉詐案件數量龐雜、基層警力資源不足、勸阻者和受害者之間的認知偏差等因素，預警勸阻工作往往難以取得顯著成效。本文通過分析當下反詐預警勸阻工作的難處，結合勸阻主體執法語言能力、勸阻話術、情景訓練技術、對話系統等要素，探究將執法語言智能訓練系統應用於反詐課程建設、反詐話術能力培訓、現場執法語言培訓的可行性，以此提高反詐預警勸阻主體業務素養、優化反詐業務工作。

關鍵詞：
電信網絡詐騙 人工智慧 預警勸阻 執法語言訓練

Applied R esearch o n t he I ntelligent T raining S ystem B ased o n L aw E nforcement Language in the Field of Anti-Fraud Early Warning and Dissuasion
Yang Yong; Taowei Guangyang; Deng Yuan

Abstract：
Owing to the non-contact and virtual nature of network crime and the lack of traditional material evidence, it is particularly difficult to investigate and crack down on telecom network fraud. Therefore, to reduce the incidence rate at the source, the ecological management of telecom network fraud crime should pay attention to the source management while crime prevention should also be the core. The prevention of electric fraud crime is divided into two parts: early warning and dissuasion. Anti-fraud early warning is mainly propaganda work, while anti-fraud dissuasion is mainly manual intervention for fraud cases that will happen or have already happened. However, due to the large number of fraud cases, the lack of police resources from the ordinary level and the differences in cognition between the dissuaders and the victims, it is often difficult to achieve significant results in early warning and dissuasion. By analyzing the current difficulties encountered in anti-fraud early warning and dissuasion, this paper explores the feasibility of applying the law enforcement language intelligent training system to the construction of anti-fraud courses, the training of antifraud skills and the on-site law enforcement language training by combining elements like the law enforcement language ability of the dissuasive body, the dissuasive speech technique, the situational training technique, the dialogue system, etc. so as to improve the quality of the work done by the anti-fraud early warning and dissuasive body in addition to optimizing the anti-fraud tasks.

Key words：
Telecom Network Fraud Artificial Intelligence Early Warning and Dissuasion Law Enforcement Language Training

DOI：
https://doi.org/10.63334/esfsmjournal.i6.2024.09.08