Automatic Creation of Knowledge Graphs from Digital Musical Document Libraries

摘要: Most of the current musicological knowledge is present in printed books and manuscripts. In last years greats efforts have been done order to digitize make available these documents form Digital Libraries. However, digital are mainly stored as raw text, with no more structure than indexes some metadata. Therefore, implicit contained text not understandable by computers cannot be processed like that. Automatic processing may help musicologists several ways, such improving navigation through a library, discovering hidden knowledge, accelerating tedious tasks, etc. To apply techniques Library, information should carefully structured semantically annotated. Information Extraction discipline computer science focused on extraction from unstructured sources. We propose method automatically extract meaningful Musical Document Libraries, using techniques. Our has two main steps. First, relevant named entities (e.g. composers, organizations, places, etc.) identified text. Second, words between syntactically analyzed understand relationship them. Finally, extracted represented machine-readable format graph, where nodes, relations edges. The resulting representation finally visualized an interactive graph. With proposed visualization, users go one document another browsing tested our subset artist biographies Grove Music Online.