A word-embedding-based steganalysis method for linguistic steganography via synonym substitution
The development of steganography technology threatens the security of privacy information in smart campus. To prevent privacy disclosure, a linguistic steganalysis method based on word embedding is proposed to detect the privacy information hidden in synonyms in the texts. With the continuous Skip-g...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2018
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/103246 http://hdl.handle.net/10220/47273 |
| _version_ | 1824454648572936192 |
|---|---|
| author | Xiang, Lingyun Yu, Jingmin Yang, Chunfang Zeng, Daojian Shen, Xiaobo |
| author2 | School of Computer Science and Engineering |
| author_facet | School of Computer Science and Engineering Xiang, Lingyun Yu, Jingmin Yang, Chunfang Zeng, Daojian Shen, Xiaobo |
| author_sort | Xiang, Lingyun |
| collection | NTU |
| description | The development of steganography technology threatens the security of privacy information in smart campus. To prevent privacy disclosure, a linguistic steganalysis method based on word embedding is proposed to detect the privacy information hidden in synonyms in the texts. With the continuous Skip-gram language model, each synonym and words in its context are represented as word embeddings, which aims to encode semantic meanings of words into low-dimensional dense vectors. The context fitness, which characterizes the suitability of a synonym by its semantic correlations with context words, is effectively estimated by their corresponding word embeddings and weighted by TF-IDF values of context words. By analyzing the differences of context fitness values of synonyms in the same synonym set and the differences of those in the cover and stego text, three features are extracted and fed into a support vector machine classifier for steganalysis task. The experimental results show that the proposed steganalysis improves the average F-value at least 4.8% over two baselines. In addition, the detection performance can be further improved by learning better word embeddings. |
| first_indexed | 2025-02-19T03:25:39Z |
| format | Journal Article |
| id | ntu-10356/103246 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2025-02-19T03:25:39Z |
| publishDate | 2018 |
| record_format | dspace |
| spelling | ntu-10356/1032462020-03-07T11:50:49Z A word-embedding-based steganalysis method for linguistic steganography via synonym substitution Xiang, Lingyun Yu, Jingmin Yang, Chunfang Zeng, Daojian Shen, Xiaobo School of Computer Science and Engineering Steganography DRNTU::Engineering::Computer science and engineering Steganalysis The development of steganography technology threatens the security of privacy information in smart campus. To prevent privacy disclosure, a linguistic steganalysis method based on word embedding is proposed to detect the privacy information hidden in synonyms in the texts. With the continuous Skip-gram language model, each synonym and words in its context are represented as word embeddings, which aims to encode semantic meanings of words into low-dimensional dense vectors. The context fitness, which characterizes the suitability of a synonym by its semantic correlations with context words, is effectively estimated by their corresponding word embeddings and weighted by TF-IDF values of context words. By analyzing the differences of context fitness values of synonyms in the same synonym set and the differences of those in the cover and stego text, three features are extracted and fed into a support vector machine classifier for steganalysis task. The experimental results show that the proposed steganalysis improves the average F-value at least 4.8% over two baselines. In addition, the detection performance can be further improved by learning better word embeddings. Published version 2018-12-28T05:59:56Z 2019-12-06T21:08:20Z 2018-12-28T05:59:56Z 2019-12-06T21:08:20Z 2018 Journal Article Xiang, L., Yu, J., Yang, C., Zeng, D., & Shen, X. (2018). A word-embedding-based steganalysis method for linguistic steganography via synonym substitution. IEEE Access, 6, 64131-64141. https://hdl.handle.net/10356/103246 http://hdl.handle.net/10220/47273 10.1109/ACCESS.2018.2878273 en IEEE Access © 2018 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. 11 p. application/pdf |
| spellingShingle | Steganography DRNTU::Engineering::Computer science and engineering Steganalysis Xiang, Lingyun Yu, Jingmin Yang, Chunfang Zeng, Daojian Shen, Xiaobo A word-embedding-based steganalysis method for linguistic steganography via synonym substitution |
| title | A word-embedding-based steganalysis method for linguistic steganography via synonym substitution |
| title_full | A word-embedding-based steganalysis method for linguistic steganography via synonym substitution |
| title_fullStr | A word-embedding-based steganalysis method for linguistic steganography via synonym substitution |
| title_full_unstemmed | A word-embedding-based steganalysis method for linguistic steganography via synonym substitution |
| title_short | A word-embedding-based steganalysis method for linguistic steganography via synonym substitution |
| title_sort | word embedding based steganalysis method for linguistic steganography via synonym substitution |
| topic | Steganography DRNTU::Engineering::Computer science and engineering Steganalysis |
| url | https://hdl.handle.net/10356/103246 http://hdl.handle.net/10220/47273 |
| work_keys_str_mv | AT xianglingyun awordembeddingbasedsteganalysismethodforlinguisticsteganographyviasynonymsubstitution AT yujingmin awordembeddingbasedsteganalysismethodforlinguisticsteganographyviasynonymsubstitution AT yangchunfang awordembeddingbasedsteganalysismethodforlinguisticsteganographyviasynonymsubstitution AT zengdaojian awordembeddingbasedsteganalysismethodforlinguisticsteganographyviasynonymsubstitution AT shenxiaobo awordembeddingbasedsteganalysismethodforlinguisticsteganographyviasynonymsubstitution AT xianglingyun wordembeddingbasedsteganalysismethodforlinguisticsteganographyviasynonymsubstitution AT yujingmin wordembeddingbasedsteganalysismethodforlinguisticsteganographyviasynonymsubstitution AT yangchunfang wordembeddingbasedsteganalysismethodforlinguisticsteganographyviasynonymsubstitution AT zengdaojian wordembeddingbasedsteganalysismethodforlinguisticsteganographyviasynonymsubstitution AT shenxiaobo wordembeddingbasedsteganalysismethodforlinguisticsteganographyviasynonymsubstitution |