基于网络文本语义的信息隐藏方法研究

The existing text steganography method is facing huge security threats and challenges, and thus there is an urgent need to find new effective method to solve these problems. As the main carriers of information dissemination in modern society, network text has many beneficial features and advantages for information hiding. So, research on network text steganography has important theoretical and practical significance. This project innovatively uses the many features of network text to study a novel information hiding method based on semantemes of network text. The main research contents of this project include: (1) We'll try to study the construction and organization strategy, mechanism and approach of the thesaurus for the use of information hiding based on the semantemes of texts, so as to establish the thesaurus for the use of efficient information hiding; (2) In order to improve the secrecy, robustness and capacity of text information hiding, we'll firstly try to study the effective synonym replace mechanism and algorithm for network text, and then to explore the new information hiding method based on the characteristics of network text; (3) To explore an effective evaluation model and method for semantic information hiding of network text, and then the improvement and perfection of the method proposed in this project will be put forward; (4) We'll finally do experimental study, so as to lay the foundation of the practical application of the research results generated by the proposed project. Our project will find out a novel and effective way for information hiding based on network text. The proposed project will also provide theoretical support for secret communication and copyright protection of network text production based on the semantic of network text, and thus enhance the research and application level of our country in related fields. The work done in this project will also enrich and develop the theory and technology system of information hiding area.

现有文本信息隐藏方法正面临巨大安全威胁和挑战，迫切需要寻找有效的新方法。作为当代信息传播的主要载体，网络文本具有诸多有利于信息隐藏的特点和优势，网络文本信息隐藏的研究具有重要的理论和实际意义。项目创新性地利用网络文本的诸多特征，研究基于网络文本语义的信息隐藏新方法，即：（1）研究面向文本语义信息隐藏的同义词库构建及组织的策略、机制和方法，建立面向高效信息隐藏的同义词库；（2）研究有效的网络文本同义词替换机制和算法，探索基于网络文本特征的信息隐藏新方法，提高信息隐藏的隐秘性、鲁棒性及容量；（3）探索网络文本语义信息隐藏的有效测评模型和方法，改进和完善项目所提出的方法；（4）开展实验研究，为项目成果的实际应用奠定基础。项目将为网络文本信息隐藏探索一种有效的新思路，为基于网络文本语义的秘密通信和网络文本知识产权保护提供理论支撑，提升我国相关领域研究和应用水平，丰富和发展信息隐藏理论技术体系。

顺应新时代网络应用的广泛普及，以项目申请书为基础，在对项目研究内容进行了具体化的前提下，本项目主要开展的研究工作有以下几点：（1）网络大数据环境下的文本信息隐藏方法研究，具体包括：设计了基于网络大数据的信息隐藏模型；建立了最优隐藏载体与秘密消息的关系模型；设计并实现了网络大数据环境下的信息隐藏系统。（2）随着社会经济的繁荣，古诗词爱好者越来越多，大量的业余爱好者将他们自己的作品发布在社交网络上，这为古诗词作为隐藏载体提供了一个极好的网络环境。基于此，本项目研究了基于唐诗宋词等格律文本智能生成的无载体信息隐藏方法，具体包括：基于格律文本自动生成的无载体信息隐藏的系统框架与模型；面向信息隐藏的深度神经网络语言模型的构建方法；文本无载体信息隐藏中机密信息编码规则和提取算法。（3）网络空间中所存在的海量的情感文本资源为秘密信息嵌入提供了新渠道，本项目开展了基于网络文本情感计算的信息隐藏方法的探索，取得了较好的效果，相关工作正在继续深入进行，可望在后续研究中取得新的研究结果。为了检验所研究的信息隐藏方法的效果，项目通过实验对所研究算法进行了分析测评。此外，在项目资助下，项目组还开展了基于语音编码的信息隐藏方法以及算法优化方面的研究。