Towards Automatic Annotation of Text Type Structure: Experiments Using an Xml-annotated Corpus and Automatic Text Classification Methods

Scientific articles exhibit a fairly conventionalized structure in terms of topic types such as background, researchTopic, method and their ordering and rhetorical interrelations. This paper describes an effort to make such structures explicit by providing a corpus of German linguistic articles with XML markup according to a text type schema defining 21 topic type categories. The corpus is further augmented with XML annotations on a grammatical level and a logical structure level. The efficiency of an automatic annotation of text type structure is explored in experiments that apply general, domain-independent automatic text classification methods to text segments and employ features from the raw text level and the corpus annotations on the grammatical level. The results indicate that some of our topic types are successfully learnable.