In this poster, a Hungarian ethnographic database containing linguistic annotation is presented. The corpus contains texts from three domains, namely, folk beliefs, t altos texts and tales. All the possible morphosyntactic analyses assigned to each word and the appropriate one selected from them (based on contextual information) are also marked. Syntactic (dependency) annotation is added semi-automatically to the corpus texts at a second phase of the processing. With the help of these enriched linguistic attributes, the texts can be semantically analyzed and clustered. The research and development team is working on a semantic search tool enabling to browse the texts on the basis of their semantic meaning. The proposed technology may result in a new approach to the ethnographic research and may open a new type of access to the databases.