英文写作学习

Abstract

• Recent studies have mostly focused on developing deep learning approaches to learn (a compact graph embedding, upon which classic clustering methods like k-means or spectral clustering algorithms are applied).
• The self-training process is jointly learned and optimized with the (graph embedding) in a unified framework,to mutually benefit both components. Experimental results compared with state-of-the-art algorithms demonstrate the superiority of our method.

Introduction

• The development of networked applications has resulted in an overwhelming number of scenarios in which data is naturally represented in graph format rather than flat-table or vector format.
• However, the complexity of (graph structure) has imposed significant challenges on these graph-related learning tasks, including graph clustering, which is one of the most popular topics.
• To solve this problem, more recent studies have resorted to deep learning techniques to learn (compact representation to exploit the rich information of both the content and structure data.)
• To achieve mutual benefit for (these two steps, a goal-directed training) framework is highly desirable. However, traditional (goal-directed training models) are mostly applied to the (classification task).

结构更新记录：
首先通过任务介绍引出论文所要研究的大背景。
接下指出该问题的关键点，给出与论文创新对应的切入点。
然后说现有前沿方法没有考虑到哪些，再引入自己所要创新和加入的那部分。也可能是两部分的结合。
最后明确给出我们需要关注的问题。以及准备构建什么样的模型。

Related Work

• (Graph clustering) has been a long-standing research topic. Early methods have taken various shallow approaches to
  (graph clustering).