Cupid is a generic schema matching algorithm that discovers mappings between schema elements based on names, data types, constraints, and schema structure. It integrates linguistic and structural matching, uses context-dependent matching for shared types, and is biased toward leaf structure where much of the schema content resides. Cupid is compared to two other schema matching systems in experiments, showing its effectiveness in matching schemas with similar structures and content. The algorithm is generic and can be applied to various data models and application domains. Cupid uses a combination of linguistic and structural similarity to compute mappings between schema elements. It first performs linguistic matching based on names, data types, and other textual descriptions, then structural matching based on the similarity of contexts or vicinities. The weighted similarity is a combination of linguistic and structural similarity. Cupid is able to overcome differences in schema element names due to normalization, is robust to different nesting of schema elements, and can disambiguate context-dependent mappings. Cupid is compared to DIKE and MOMIS in experiments, showing that it performs better in matching schemas with similar structures and content. Cupid is able to infer relationships between schemas, such as the correspondence of a single table in one schema to the join of two tables in another. Cupid is also able to handle referential constraints, which are important for schema matching. The algorithm is generic and can be applied to various data models and application domains. Cupid is able to handle schemas with shared types and referential constraints, which are important for schema matching. Cupid is able to handle schemas with complex structures and nested elements, which are common in real-world applications. Cupid is able to handle schemas with different data types and structures, which are common in real-world applications. Cupid is able to handle schemas with different data models, which are common in real-world applications. Cupid is able to handle schemas with different data models, which are common in real-world applications. Cupid is able to handle schemas with different data models, which are common in real-world applications.Cupid is a generic schema matching algorithm that discovers mappings between schema elements based on names, data types, constraints, and schema structure. It integrates linguistic and structural matching, uses context-dependent matching for shared types, and is biased toward leaf structure where much of the schema content resides. Cupid is compared to two other schema matching systems in experiments, showing its effectiveness in matching schemas with similar structures and content. The algorithm is generic and can be applied to various data models and application domains. Cupid uses a combination of linguistic and structural similarity to compute mappings between schema elements. It first performs linguistic matching based on names, data types, and other textual descriptions, then structural matching based on the similarity of contexts or vicinities. The weighted similarity is a combination of linguistic and structural similarity. Cupid is able to overcome differences in schema element names due to normalization, is robust to different nesting of schema elements, and can disambiguate context-dependent mappings. Cupid is compared to DIKE and MOMIS in experiments, showing that it performs better in matching schemas with similar structures and content. Cupid is able to infer relationships between schemas, such as the correspondence of a single table in one schema to the join of two tables in another. Cupid is also able to handle referential constraints, which are important for schema matching. The algorithm is generic and can be applied to various data models and application domains. Cupid is able to handle schemas with shared types and referential constraints, which are important for schema matching. Cupid is able to handle schemas with complex structures and nested elements, which are common in real-world applications. Cupid is able to handle schemas with different data types and structures, which are common in real-world applications. Cupid is able to handle schemas with different data models, which are common in real-world applications. Cupid is able to handle schemas with different data models, which are common in real-world applications. Cupid is able to handle schemas with different data models, which are common in real-world applications.