大数据渗透到社会生活的方方面面，从互联网公司利用收集到的用户数据向用户推荐个性化信息、商品、生活服务方案，到新冠疫情，中国流调部门利用大数据监测感染人群及密集接触人员的流动信息来辅助防疫部门进行防疫部署。另一方面，技术的发展给犯罪分子以可乘之机，犯罪逐渐呈现远距离化、手段多样化、高科技化趋势。

当下，侦查模式主要以传统模式为主，倾向于事后侦查。即从已形成的犯罪结果回溯，通过搜集证据、审讯嫌疑人来证明犯罪事实、还原犯罪经过。侦破案件需要的周期长、警力多，侦查过程依赖经验与口供，侦查工作开展效率低下。传统侦查模式已经无法适应犯罪现状。

因此，侦查模式的转型是时代的选择。大数据技术的发展、数据资源的战略定位以及犯罪的态势，为数据驱动型侦查模式的实施提供可行性与必要性。数据驱动的侦查模式应当注重预测性、主动性、全景性，以物质交换理论、相关性理论为基础，在侦查理念，侦查思维，侦查手段等方面转型。

本文共分为四个部分：第一部分介绍了大数据的概念、内容及特点，旨在说明大数据本身的特点和功能及其发展现状为数据驱动侦查提供了技术基础。同时，对大数据应用于侦查工作的优劣进行了分析，旨在更为全面地认识和理解数据驱动型侦查模式，避免“唯线上论”，提倡“线上与线下相结合”的侦查模式。第二部分通过对传统侦查模式特点、思维、手段的解构，证明了传统侦查模式已经不能适应如今的司法理念、犯罪现状、侦查效益和犯罪预防策略。因此，传统侦查模式需要转型。第三部分，分析了侦查模式转型的必要性，说明侦查模式转型是时代的选择，是侦查工作突破瓶颈的必要手段。中国对于大数据发展的态度及政策让侦查模式的转型具有可行性。第四部分对于数据驱动型侦查模式进行了全方位的分析。实践层面，通过国内外侦查模式转型实例说明了数据驱动型侦查模式是传统侦查模式转型的目标。从理论层面，将数据驱动型侦查模式与其他主流侦查模式对比，从控制论、信息论、博弈论、法律经济学视角说明传统侦查模式应当转型为数据驱动型侦查模式。本文最后，对于数据驱动型侦查模式的实现路径进行了建议。

Big data permeates every aspect of social life, from Internet companies using collected user data to recommend personalized information, goods, and life service solutions to the new crown epidemic China's flow investigation department uses big data to monitor the flow information of infected people and intensive contacts to assist the epidemic prevention department with epidemic prevention arrangements. The era of big data has quietly arrived. At the same time, it is the contradiction that the current investigation model is lagging behind, inefficiency and lack of manpower cannot adapt to the situation of diverse, intelligent and dynamic crime.

The current investigation model is mainly based on the traditional model, tending to post facto investigation. That is, from the results of the formed crime back, through the collection of evidence, interrogation of suspects to prove the facts of the crime, restore the crime. Such an investigation mode has a long period of time and requires more police force to solve cases. In the future, crime presents the development trend of non-violence and diversification, with characteristics of intelligence, organization and dynamics. The traditional investigation model is obviously unable to adapt to the development of crime and thus effectively combat crime.

The transformation of the investigation model is the choice of the times. The development of big data technology, the national determination to develop big data and the crime situation, provide the feasibility and necessity of data-driven investigation mode. Data-driven investigation model should focus on predictive, preventive, panoramic, based on the theory of material exchange theory, relevance theory, in the investigation of initiation, entry, detection, thinking, means and other aspects of the traditional investigation model to break through.

This paper is divided into five parts: the first part introduces the concept, content and characteristics of big data, aiming to explain the characteristics and functions of big data itself and its development status provides the technical basis for data-driven investigation. The second part analyzes the advantages and contradictions brought by the application of big data in the investigation work, aiming at a more comprehensive understanding of the data-driven investigation model, avoiding the "online only theory" and advocating the "online and offline combination" of the investigation model. The third part introduces the limitations of the traditional investigation model. Through the opportunity of transformation combined with the limitations of the traditional investigation model, the fourth part proves the necessity of the transformation of the investigation model, indicating that the transformation of the investigation model is the choice of the times, is a necessary means to break through the bottleneck of the investigation work. And the world and China's attitude and policies on big data prove that the transformation of investigation mode is feasible. The fifth part is a comprehensive analysis of the data-driven investigation model, mainly from its content and characteristics, combined with application examples to show that the data-driven investigation model has been successful experience. At the same time, the challenges encountered in the transformation of the investigative model are suggested path options, providing the way of transformation, and finally, the goals and directions to be achieved by the data-driven investigative model are proposed.