传统的静态检测恶意JavaScript代码方法十分依赖于已有的恶意代码特征,无法有效提取混淆恶意代码特征,导致检测混淆恶意JavaScript代码的精确率低。针对该问题提出基于双向长短期记忆网络(Bidirectional Long Short-term Memory, Bi-LS...传统的静态检测恶意JavaScript代码方法十分依赖于已有的恶意代码特征,无法有效提取混淆恶意代码特征,导致检测混淆恶意JavaScript代码的精确率低。针对该问题提出基于双向长短期记忆网络(Bidirectional Long Short-term Memory, Bi-LSTM)的恶意代码检测模型。通过抽象语法树将JavaScript代码转化为句法单元序列,通过Doc2Vec算法将句法单元序列用分布式向量表示,将句向量矩阵送入Bi-LSTM模型进行检测。实验结果表明,该方法对于混淆恶意JavaScript代码具有良好的检测效果且检测效率高,准确率为97.03%,召回率为97.10%。展开更多
Rivers are progressively being exposed to increased anthropogenic pollution stresses that are undermining their designated uses and affecting sensitive coastal areas. In this study, three adjacent eastern Mediterranea...Rivers are progressively being exposed to increased anthropogenic pollution stresses that are undermining their designated uses and affecting sensitive coastal areas. In this study, three adjacent eastern Mediterranean coastal rivers, Ibrahim, Kaleb and Beirut, were evaluated. Water quality samples were collected in dry and wet seasons from different sampling sites along the river from the source to the outlet which represent a gradient of increased urbanization. The spatiotemporal variability of the physio-chemical properties, heavy metals (Zn, Pb, Cu, Cr, and Cd) and organic matter (DOC) were statistically analyzed to better understand the contribution of point and nonpoint pollution sources. The three rivers (Beirut, Kaleb and Ibrahim) show a similar behavior in calcium and carbonate alkalinity due to the carbonate mineral weathering effect, so they are of calcium bicarbonate type due to their calcareous geological nature. The speciation of anions was affected by temporal variation. Moreover, it is obvious that the Beirut River has a different behavioral characteristic where the water is a sulfate type water with a preferable metal-OM complexation mainly with lead, zinc and copper, whereas Kaleb and Ibrahim are considered to be of a nitrate phosphate type with a preferable metal inorganic complexation, especially copper, that has a consistent behavior in both types of waters. This difference is attributed to the urbanization effect highly impacting the Beirut River.展开更多
文摘传统的静态检测恶意JavaScript代码方法十分依赖于已有的恶意代码特征,无法有效提取混淆恶意代码特征,导致检测混淆恶意JavaScript代码的精确率低。针对该问题提出基于双向长短期记忆网络(Bidirectional Long Short-term Memory, Bi-LSTM)的恶意代码检测模型。通过抽象语法树将JavaScript代码转化为句法单元序列,通过Doc2Vec算法将句法单元序列用分布式向量表示,将句向量矩阵送入Bi-LSTM模型进行检测。实验结果表明,该方法对于混淆恶意JavaScript代码具有良好的检测效果且检测效率高,准确率为97.03%,召回率为97.10%。
文摘Rivers are progressively being exposed to increased anthropogenic pollution stresses that are undermining their designated uses and affecting sensitive coastal areas. In this study, three adjacent eastern Mediterranean coastal rivers, Ibrahim, Kaleb and Beirut, were evaluated. Water quality samples were collected in dry and wet seasons from different sampling sites along the river from the source to the outlet which represent a gradient of increased urbanization. The spatiotemporal variability of the physio-chemical properties, heavy metals (Zn, Pb, Cu, Cr, and Cd) and organic matter (DOC) were statistically analyzed to better understand the contribution of point and nonpoint pollution sources. The three rivers (Beirut, Kaleb and Ibrahim) show a similar behavior in calcium and carbonate alkalinity due to the carbonate mineral weathering effect, so they are of calcium bicarbonate type due to their calcareous geological nature. The speciation of anions was affected by temporal variation. Moreover, it is obvious that the Beirut River has a different behavioral characteristic where the water is a sulfate type water with a preferable metal-OM complexation mainly with lead, zinc and copper, whereas Kaleb and Ibrahim are considered to be of a nitrate phosphate type with a preferable metal inorganic complexation, especially copper, that has a consistent behavior in both types of waters. This difference is attributed to the urbanization effect highly impacting the Beirut River.
