Self-adaptive mobile web service (MWS) discovery in Smart Systems depends on heterogeneous context-sensitive attributes that are composed dynamically to satisfy the user’s MWS request despite the constantly changing environment. Several ontologies are developed to deal with the issues of heterogeneous knowledge representation in smart systems. Unfortunately, these ontologies are mostly inexpressive, unextendible, constricted, and slightly fragmented due to a lack of socially focused ontology development procedures. In this paper, a context ontology for self-adaptive MWS discovery in smart systems is proposed to provide a better representation of the heterogeneous context for smart systems and facilitate the delivery of MWS. The lightweight unified process for ontology building (UPON Lite) is adopted for ontology development. The ontology is experimentally evaluated in protégé using real-world smart systems from the healthcare and agriculture domain. Consequently, the ontology was found to be more expressive, extensible, and support self-adaptive MWS discovery in Smart Systems.