Linguistic features play an important role in the transition from traditional citation analysis to citation content analysis (CCA). In this paper, a novel language feature, Reporting Marker (RM), is introduced into CCA and acts as a key to open the black box of information structure of citation content. Through discourse analysis performed on a small-scale corpus, we clarify the mapping relationship between information pattern and language structure, and use the Language Technology Platform (LTP) to identify that relationship. A framework is constructed to realize CCA in three dimensions: citation network, knowledge, and evaluation. It is likely to contribute to the improvement of information retrieval and knowledge mining from academic documents.