Fig. 1. A connected partition example of the Event hierarchy of the SN.

Fig. 2. Metaschema hierarchy corresponding to the partition of the Event hierarchy of Fig. 1.

Fig. 3. Interpretation of the definition of meta-child-of.

Fig. 4. Top-down consensus metaschema hierarchy.

Fig. 5. The abstract Semantic Network and input metaschemas.

Fig. 6. The flow chart of consolidation numbered steps.

Fig. 7. Consolidating and deleting corresponding MSTs.

Fig. 8. Constructing and S″.

Fig. 9. Constructing .

Fig. 10. Consolidated metaschema.

Fig. 11. Bottom-up consensus metaschema hierarchy.

Fig. 12. Comparison of consensus metaschemas.

Fig. 13. The consolidation of corresponding MSTs.

Fig. 14. The partial consolidated metaschema with corresponding MSTs.

Fig. 15. Auxiliary metaschemas , and expanded Semantic Network SN′.

Fig. 16.  and for Entity.

Fig. 17. Processing for Entity.

Fig. 18. Consolidated metaschema hierarchy.

Fig. 19. Consolidated metaschema with relationships.

Number of MSTs each expert chose in the top-down study

Inter-participant agreement matrix; average = 16.76

Number of semantic types marked by at least N participants

Number of MSTs each expert chose in the bottom-up study

Number of MSTs chosen by at least N participants

Inter-participant agreement matrix; average = 14.41

Identical MSTs in both metaschemas

Similar MSTs in both metaschemas

Refinements in bottom-up consensus metaschema

Values for the structural measures for the two consensus metaschemas

Weight distribution of the top-down and bottom-up metaschemas

Relationship number codes

Comparing experts’ metaschemas of both studies

Comparison of the structural measures for the top-down, bottom-up and consolidated metaschemas