Ontology matching can be defined as the process of discovering similarities between two ontologies, and it can be processed by exploiting a number of different techniques. To provide a common conceptual basis of ontology matching for the Semantic Web, researchers have started to develop classifications to distinguish ontologies. The most significant one is the classification proposed by Shvaiko and Euzenat. Their approach is to compare different existing ontology mediation systems as well as to design a schemabased matching system. In our findings, the above classifications contain some improper identifications and vague categories. There are three insufficient elementary matching techniques out of ten matching techniques: the language-based matching, the repository of structures, and the upper level formal ontology techniques. The language-based matching technique is normally performed prior to string-based technique and has no direct engagement in the actual similarity computation between two ontologies. The repository of structures technique is a dynamic approach used to compare fragments of two ontologies and to eliminate the dissimilar portions. It may be regarded as the follow-up step. The upper level formal ontologies technique is an approach that uses external source of common knowledge in the form of ontology. There is insufficient evidence that specifies the input and design guidelines. This paper therefore proposes a design and input-specific classification framework of ontology matching techniques to address the above problems based on the findings of the literature survey. The proposed framework consists of the layers: executive approach layer, basic technique layer and input layer. The executive approach layer identifies heuristic, probabilistic reasoning and semantic reasoning to execute the identified seven elementary ontology matching techniques. To provide a guideline between the executive and input layers, the basic technique layer consists of string-based, linguistic resources, constraint-based, alignment reuse, graphbased, taxonomy-based and model-based matching techniques. The input layer is classified into elementary input and structural input to sum up the characteristics of the actual inputs.