As the number and adoption of social networking sites (SNSs) supporting business representation in the form of business pages continues to escalate, more scalable and robust mechanisms for integrating data from different networks in order to serve the special purposes need to be envisaged. An important concern of such SNS data integration is the platform dependencies that different networks impose in collecting, organizing, and presenting the business information hosted on their servers. In this esteem, this paper deals with overcoming the challenge of different business categorization schemes being varyingly used by the existing SNSs. In doing so, we present a content-oriented approach for determining the business category on the basis of a semantic analysis of the textual information available in the business profile. The approach has been operationalized in the CoDiT (Company Discovery Tool) which is a web-based tool to facilitate the integrated business page search over multiple SNSs.