Manfred Stede

Lexicon-based methods for sentiment analysis

We present a lexicon-based approach to extracting sentiment from text. The Semantic Orientation CALculator (SO-CAL) uses dictionaries of words annotated with their semantic orientation (polarity and strength), and incorporates intensification and negation. SO-CAL is applied to the polarity classification task, the process of assigning a positive or negative label to a text that captures the texts opinion towards its main subject matter. We show that SO-CALs performance is consistent across domains and in completely unseen data. Additionally, we describe the process of dictionary creation, and our use of Mechanical Turk to check dictionaries for consistency and reliability.

The Potsdam commentary corpus

A corpus of German newspaper commentaries has been assembled and annotated with different information (and currently, to different degrees): part-of-speech, syntax, rhetorical structure, connectives, co-reference, and information structure. The paper explains the design decisions taken in the annotations, and describes a number of applications using this corpus with its multi-layer annotation.

Customizing RST for the Automatic Production of Technical Manuals

Rhetorical Structure Theory (RST) has emerged as a promising candidate for text representation in NLG. We investigated the applicability of RST in the automatic production of multilingual technical manuals. Starting from a domain knowledge base, we construct an RST-tree for a particular manual section, which is then converted to a set of sentence plans. These plans serve as input to sentence generators that produce the final text. In this paper, we report first on a number of open questions regarding general aspects of RST. Arguing that the original set of RST relations is not specific enough for practical generation purposes, we suggest a number of new relations that we found useful in our domain. After briefly examining the stage of RST tree construction, we then outline a procedure for converting RST trees to a sequence of sentence plans.

Ma(r)king concessions in English and German

Natural language generation aims at automatically verbalizing a “deep” representation of content, so that coherent and cohesive text originates. To produce such cohesive discourse, it is important to signal many of the relations holding between text segments to the reader by means of cue words, which we call discourse markers. Current generation systems usually do this in a simplistic way, e.g., by using one marker per relation. In reality, however, language offers a wide range of markers from which informed choices should be made. This paper suggests a method for equipping generators with the knowledge to select the most appropriate discourse marker from a set of candidate expressions. We concentrate on one area of discourse relations, the CONCESSION family, and identify its underlying semantics and pragmatics. On the basis of extensive corpus studies, we propose a new classification of CONCESSION markers in English and German, and then suggest a generation model for producing bilingual text that can inc...

Genre-Based Paragraph Classification for Sentiment Analysis

We present a taxonomy and classification system for distinguishing between different types of paragraphs in movie reviews: formal vs. functional paragraphs and, within the latter, between description and comment. The classification is used for sentiment extraction, achieving improvement over a baseline without paragraph classification.

Joint prediction in MST-style discourse parsing for argumentation mining

We introduce a new approach to argumentation mining that we applied to a parallel German/English corpus of short texts annotated with argumentation structure. We focus on structure prediction, which we break into a number of subtasks: relation identification, central claim identification, role classification, and function classification. Our new model jointly predicts different aspects of the structure by combining the different subtask predictions in the edge weights of an evidence graph; we then apply a standard MST decoding algorithm. This model not only outperforms two reasonable baselines and two datadriven models of global argument structure for the difficult subtask of relation identification, but also improves the results for central claim identification and function classification and it compares favorably to a complex mstparser pipeline.

Generating multilingual documents from a knowledge base: the TECHDOC project

TECHDOC is an implemented system demonstrating the feasibility of generating multilingual technical documents on the basis of a language-independent knowledge base. Its application domain is user and maintenance instructions, which are produced from underlying plan structures representing the activities, the participating objects with their properties, relations, and so on. This paper gives a brief outline of the system architecture and discusses some recent developments in the project: the addition of actual event simulation in the KB, steps towards a document authoring tool, and a multimodal user interface.

Lexicalization in natural language generation: a survey

In natural language generation, a meaning representation of some kind is successively transformed into a sentence or a text. Naturally, a central subtask of this problem is the choice of words, orlexicalization. In this paper, we propose four major issues that determine how a generator tackles lexicalization, and survey the contributions that researchers have made to them. Open problems are identified, and a possible direction for future research is sketched.

Lexical semantics and knowledge representation in multilingual text generation

List of Figures. List of Tables.Preface. 1. Introduction. 2. Lexicalization in NLG. 3. Classifying Lexical Variation. 4. Modelling the Domain. 5. Levels of Representation: SITSPEC and SEMSPEC. 6. Representing the Meaning of Words. 7. Verb Alternations and Extensions. 8. A System Architecture for Multilingual Generation. 9. Generating Paraphrases. 10. From Sentences to Text. 11. Summary and Conclusions. References. Index.

By all these lovely tokens... Merging Conflicting Tokenizations

Given the contemporary trend to modular NLP architectures and multiple annotation frameworks, the existence of concurrent tokenizations of the same text represents a pervasive problem in everyday’s NLP practice and poses a non-trivial theoretical problem to the integration of linguistic annotations and their interpretability in general. This paper describes a solution for integrating different tokenizations using a standoff XML format, and discusses the consequences from a corpus-linguistic perspective.