Artificial Intelligence Review - In recent times, there have been attempts to leverage artificial intelligence (AI) techniques in a broad range of cyber security applications. Therefore, this paper... 相似文献
The Journal of Supercomputing - High-precision point cloud maps have drawn increasing attention due to their wide range of applications. In recent decades, point cloud maps are normally generated... 相似文献
The traditional emotion–cause extraction task needs to give the exact emotion annotation contained in the document before extracting the cause. Different from this, the emotion–cause pair extraction (ECPE) task, which aims to extract emotion–cause pairs with causal relationships directly from the document, is a task proposed in the natural language processing field recently. At present, the task of ECPE is divided into two steps: emotion annotations and cause clause extraction, emotion–cause clause pair combining and filtering. In this article, we optimize these two steps. On the one hand, in the first step of ECPE, a mutual assistance single-task model proposed by us is used to replace the original multi-task model. On the other hand, the position information of the clause is added as an additional feature in the second step of ECPE. Furthermore, based on different levels of semantic features, we design three filtering models and explore their performance on ECPE tasks. The experimental results on the benchmark corpus show that our approach can make the ECPE task achieve better performance. Compared with the referenced method, F1-score is increased by 5.3%. Moreover, these optimization strategies improve the subtasks contained in ECPE to varying degrees.
