Situation entity types: automatic classification of clause-level aspect

Abstract

This paper describes the first robust approach to automatically labeling clauses with their situation entity type (Smith, 2003), capturing aspectual phenomena at the clause level which are relevant for interpreting both semantics at the clause level and discourse structure. Previous work on this task used a small data set from a limited domain, and relied mainly on words as features, an approach which is impractical in larger settings. We provide a new corpus of texts from 13 genres (40,000 clauses) annotated with situation entity types. We show that our sequence labeling approach using distributional information in the form of Brown clusters, as well as syntactic-semantic features targeted to the task, is robust across genres, reaching accuracies of up to 76%.

Semantic clause types and modality as features for argument analysis1

“…Using automatic classifiers to assign situation entity [9] and modal sense labels [15,16], the contribution of these phenomena to argument structure parsing can be examined on a broader scale.…”

Classifying Semantic Clause Types: Modeling Context and Genre Characteristics with Recurrent Neural Networks and Attention

“…The features used by Friedrich et al (2016) cover a variety of linguistic features -such as tense, voice, number, POS, semantic clusters -some of which we expect to be encoded in pre-trained embeddings, while others will emerge through model training. We start with pre-trained embeddings for both English and German, because this leads to better results than random ini- tialization.…”

“…For the English dataset, we use the same testtrain split as Friedrich et al (2016). 8 The German dataset was split into training and testing with a balanced distribution of genres (as is the case for the English dataset).…”

“…Recent work on automatic identification of SE types relies on feature-based classifiers for English that have been successfully applied to various textual genres (Friedrich et al, 2016), and also show that a sequence labeling approach that models contextual clause labels yields improved classification performance.…”

“…Given the highly-engineered feature sets used for SE classification so far (Friedrich et al, 2016), porting such classifiers to other languages is a non-trivial issue. We test the portability of our system by applying it to German.…”