This paper aims to show how to develop a knowledge graph as linguistic knowledge-base for linguistic artificial intelligence applications. Firstly, we investigate the previous researches and developments of knowledge graphs which have been developed so far and are the extensions of lexicon and ontology which are well-known forms of linguistic knowledge bases. Secondly, we describe how to develop a knowledge graph in the domain of affective disorder as a case study based on expert knowledge data and general text data. In the process of doing this, we would like to show how to collect language resources from the Web which have popular influences on people and try to make formalisation of representation way of linguistic knowledge in the domain of psychology. Finally, we show the application of using a knowledge graph as a semantic knowledge-base. Overall, we would say that a linguistic knowledge graph still would be one of the knowledge powers which will be linguistically explainable by humans in the world of digital data analysis floods.