nach oben

2010 | Buch

Kapitel lesen Erstes Kapitel lesen

Controlled Natural Language

Workshop on Controlled Natural Language, CNL 2009, Marettimo Island, Italy, June 8-10, 2009. Revised Papers

herausgegeben von: Norbert E. Fuchs

Verlag: Springer Berlin Heidelberg

Buchreihe : Lecture Notes in Computer Science

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Einloggen, um Zugang zu erhalten

Über dieses Buch

Controlled natural languages (CNLs) are subsets of natural languages, obtained by - stricting the grammar and vocabulary in order to reduce or eliminate ambiguity and complexity. Traditionally, controlled languagesfall into two major types: those that - prove readability for human readers, and those that enable reliable automatic semantic analysis of the language. [. . . ] The second type of languages has a formal logical basis, i. e. they have a formal syntax and semantics, and can be mapped to an existing formal language, such as ?rst-order logic. Thus, those languages can be used as knowledge representation languages, and writing of those languages is supported by fully au- matic consistency and redundancy checks, query answering, etc. Wikipedia Variouscontrollednatural languagesof the second type have been developedby a n- ber of organizations, and have been used in many different application domains, most recently within the Semantic Web. The workshop CNL 2009 was dedicated to discussing the similarities and the d- ferences of existing controlled natural languages of the second type, possible impro- ments to these languages, relations to other knowledge representation languages, tool support, existing and future applications, and further topics of interest.

Inhaltsverzeichnis

Frontmatter

Language Aspects

An Evaluation Framework for Controlled Natural Languages

Abstract

This paper presents a general framework called ontographs that relies on a graphical notation and enables the tool-independent and reliable evaluation of human understandability of knowledge representation languages. An experiment with 64 participants is presented that applies this framework and compares a controlled natural language to a common formal language. The results show that the controlled natural language is easier to understand, needs less learning time, and is more accepted by its users.

Tobias Kuhn

Rhetorical Compositions for Controlled Natural Languages

Abstract

Logic-based controlled natural languages usually provide some facility for compositional representation, minimally including sentence level coordination and sometimes subordination. These forms of compositional representation are useful for expressing short passages of a few sentences, but compared to natural language they offer only a limited range of expression, and they are unwieldy for expressing longer passages. This paper describes a method for bringing more naturalness to controlled natural languages. This approach defines a model for representing compositional structures found in natural language, such as Antithesis, Concession, and Unless in a way that is both rhetorically expressive and logically reducible. This is demonstrated through a series of examples using both natural and controlled natural language. The paper then describes a set of intercompositional rules of inference that can be used to reason about the interrelationships between compositional structures, facilitating the discovery and assessment of supportive and conflicting relationships among them. This ability to represent and discover interrelationships among discourse representation structures could be useful for developing applications that must not only follow a set of rules of inference, but reason about the rules themselves.

Andrew Potter

Anaphora Resolution Involving Interactive Knowledge Acquisition

Abstract

Anaphora resolution in current computer-processable controlled natural languages relies mainly on syntactic information, accessibility constraints and the distance of the anaphoric expression to its antecedent. This design decision has the advantage that a text can be processed automatically without any additional ontological knowledge, but it has the disadvantage that the author is severely restricted in using anaphoric expressions while writing a text. I will argue that we can allow for a wider range of anaphoric expressions whose resolution requires inference-supporting knowledge, if we consider the anaphora resolution process as an interactive knowledge acquisition process in those cases where no suitable noun phrase antecedent can be found. In particular, I will focus on definite descriptions that stand in a synonymy, subclass/superclass or part-whole relation to their noun phrase antecedent, and show how the original anaphora resolution algorithm of PENG Light can be extended in a systematic way in order to take care of these bridging definite descriptions. The solution to this problem also sheds some light on the adequate treatment of part-whole relations in a controlled natural language context.

Rolf Schwitter

Talking Rabbit: A User Evaluation of Sentence Production

Abstract

Rabbit is a controlled natural language (CNL) designed to aid experts in understanding and authoring domain ontologies. There are three broad types of Rabbit sentence: declarations, axioms and import statements. This paper evaluates the ease with which domain experts without any prior ontology development experience can author declarations and axiom sentences in Rabbit. Participants were asked to author Rabbit sentences about an artificial domain (Planet Zog). The most common error observed was the omission of the every keyword at the beginning of sentences. Another common error was to confuse instance and subclass declarations. Suggested improvements include changes to the Rabbit syntax as well modifications to a CNL authoring environment.

Paula Engelbrecht, Glen Hart, Catherine Dolbear

Naturalness vs. Predictability: A Key Debate in Controlled Languages

Abstract

In this paper we describe two quite different philosophies used in developing controlled languages (CLs): A "naturalist" approach, in which CL interpretation is treated as a simpler form of full natural language processing; and a "formalist" approach, in which the CL interpretation is “deterministic” (context insensitive) and the CL is viewed more as an English-like formal specification language. Despite the philosophical and practical differences, we suggest that a synthesis can be made in which a deterministic core is embedded in a naturalist CL, and illustrate this with our own controlled language CPL.

In the second part of this paper we present a fictitious debate between an ardent “naturalist” and an ardent “formalist”, each arguing their respective positions, to illustrate the benefits and tradeoffs of these different philosophies in an accessible way.

Peter Clark, William R. Murray, Phil Harrison, John Thompson

Implementing Controlled Languages in GF

Abstract

This paper introduces GF, Grammatical Framework, as a tool for implementing controlled languages. GF provides a high-level grammar formalism and a resource grammar library that make it easy to write grammars that cover similar fragments in several natural languages at the same time. Authoring help tools and automatic translation are provided for all grammars. As an example, a grammar of Attempto Controlled English is implemented and then ported to Finnish, French, German, Italian and Swedish.

Krasimir Angelov, Aarne Ranta

Polysemy in Controlled Natural Language Texts

Abstract

Computational semantics and logic-based controlled natural languages (CNL) do not address systematically the word sense disambiguation problem of content words, i.e., they tend to interpret only some functional words that are crucial for construction of discourse representation structures. We show that micro-ontologies and multi-word units allow integration of the rich and polysemous multi-domain background knowledge into CNL thus providing interpretation for the content words. The proposed approach is demonstrated by extending the Attempto Controlled English (ACE) with polysemous and procedural constructs resulting in a more natural CNL named PAO covering narrative multi-domain texts.

Normunds Gruzitis, Guntis Barzdins

Economical Discourse Representation Theory

Abstract

First-order logic (FOL) is undecidable — that is, no algorithm exists that can decide whether a formula of FOL is valid or not. However, there are various fragments of FOL that are known to be decidable. FO², the two-variable fragment of FOL, is one of such languages [1,2]. FO² is a first-order language where formulas have maximally two variables, no function symbols, but possibly do have equality. FO² has the finite model property [1], which means that if a formula of FO² is satisfiable, it is satisfiable in a finite model.

Johan Bos

Controlled English Ontology-Based Data Access

Abstract

As it is well-known, querying and managing structured data in natural language is a challenging task due to its ambiguity (syntactic and semantic) and its expressiveness. On the other hand, querying, e.g., a relational database or an ontology-based data access system is a well-defined and unambigous task, namely, the task of evaluating a formal query (e.g., an SQL query) of a limited expressiveness over such database. However these formal query languages may be difficult to learn and use for the casual user and ambiguity may compromise the interface. To bridge this gap, the use of controlled language interfaces has been proposed. As a measure of their efficiency for data access, we propose to consider data complexity, which is the complexity of query evaluation measured in the size of the data. We study a familiy of controlled languages that express several fragments of OWL, ranging from tractable (LogSpace and PTime) to intractable (coNP-hard) in data complexity, singling out which constructs give rise to each computational property.

Camilo Thorne, Diego Calvanese

SBVR’s Approach to Controlled Natural Language

Abstract

The “Semantics of Business Vocabulary and Business Rules” (SBVR 1.0) is one of the initial specifications in the OMG’s family of business-focused specifications. SBVR covers two aspects: Vocabulary (natural language ontology) and Rules (elements of guidance that govern actions). However, SBVR does not standardize any particular language for expressing vocabularies and rules. Instead, SBVR uses ‘semantic formulation’, which is a way of describing the semantic structure of statements and definitions. This approach of specifying structures of meaning, with its sound theoretical foundation of formal logic, provides a formal, language-independent means for capturing the semantics of a community’s body of shared meanings. By taking this approach, SBVR can support multiple forms of representation.

Silvie Spreeuwenberg, Keri Anderson Healy

Tools and Applications

The Naproche Project Controlled Natural Language Proof Checking of Mathematical Texts

Abstract

This paper discusses the semi-formal language of mathematics and presents the Naproche CNL, a controlled natural language for mathematical authoring. Proof Representation Structures, an adaptation of Discourse Representation Structures, are used to represent the semantics of texts written in the Naproche CNL. We discuss how the Naproche CNL can be used in formal mathematics, and present our prototypical Naproche system, a computer program for parsing texts in the Naproche CNL and checking the proofs in them for logical correctness.

Marcos Cramer, Bernhard Fisseni, Peter Koepke, Daniel Kühlwein, Bernhard Schröder, Jip Veldman

On Designing Controlled Natural Languages for Semantic Annotation

Abstract

Manual semantic annotation is a complex and arduous task both time-consuming and costly often requiring specialist annotators. (Semi)-automatic annotation tools attempt to ease this process by detecting instances of classes within text and relationships between instances, however their usage often requires knowledge of Natural Language Processing(NLP) or formal ontological descriptions. This challenges researchers to develop user-friendly annotation environments within the knowledge acquisition process. Controlled Natural Languages (CNL)s offer an incentive to the novice user to annotate, while simultaneously authoring, his/her respective documents in a user-friendly manner, yet shielding him/her from the underlying complex knowledge representation formalisms. CNLs have already been successfully applied within the context of ontology authoring, yet very little research has focused on CNLs for semantic annotation. We describe the design and implementation of two approaches to user friendly semantic annotation, based on Controlled Language for Information Extraction tools, which permit non-expert users to semi-automatically both author and annotate meeting minutes and status reports using controlled natural language.

Brian Davis, Pradeep Dantuluri, Laura Dragan, Siegfried Handschuh, Hamish Cunningham

Development of a Controlled Natural Language Interface for Semantic MediaWiki

Abstract

Semantic wikis support the collaborative creation, editing and utilization of semantically-enriched content, and they may therefore be well-suited to addressing problems associated with the limited availability of high-quality online semantic content. Unfortunately, however, many popular semantic wikis, such as Semantic MediaWiki (SMW), are not sufficiently expressive to support full-scale ontology authoring. Furthermore, the grounding of the Semantic Web in formal logic makes both the comprehension and production of ontological content difficult for many end-users. In order to address these issues, the expressivity of SMW was extended using a combination of semantic templates and a Web Ontology Language (OWL) meta-model. Semantic templates were also used to provide an ontology verbalization capability for SMW using the Rabbit Controlled Natural Language (CNL). The resulting system demonstrates how CNL interfaces can be implemented on top of SMW. The proposed solution introduces no changes to the underlying functionality of the SMW system, and the use of semantic templates as an ontology verbalization solution means that end-users can exploit all the usual features of conventional wiki systems to collaboratively create new CNL verbalization capabilities.

Paul R. Smart, Jie Bao, Dave Braines, Nigel R. Shadbolt

A Controlled Language for the Specification of Contracts

Abstract

Controlled natural languages have been used to enable the direct translation from natural language specifications into a formal description. In this paper we make a case for such an approach to write contracts, and translating into a temporal deontic logic. Combining both temporal behaviour and deontic behaviour is challenging both from a natural language and a formal logic perspective. We present both a logic and a controlled natural language and outline how the two can be linked.

Gordon J. Pace, Michael Rosner

Rabbit to OWL: Ontology Authoring with a CNL-Based Tool

Abstract

Recent work on ontology engineering has seen the adoption of controlled natural languages to ease the process of ontology authoring. However, CNL-based tools still require good knowledge engineering skills to be used efficiently. In this paper presents ROO, an ontology authoring tool that has been designed to cater for the needs of domain experts with little or no ontology engineering experience. ROO combines a CNL-based interface with appropriate tool support based on an ontology construction methodology. We focus on how this tool support is provided in ROO by using and implementing novel aspects of the Rabbit controlled natural language and we refer to an evaluation study that provides empirical evidence in support of using CNL-based techniques to assist ontology authors.

Ronald Denaux, Vania Dimitrova, Anthony G. Cohn, Catherine Dolbear, Glen Hart

Writing Clinical Practice Guidelines in Controlled Natural Language

Abstract

Clinicians could benefit from decision support systems incorporating the knowledge contained in clinical practice guidelines. However, the unstructured form of these guidelines makes them unsuitable for formal representation. To address this challenge we translated a complete set of pediatric guideline recommendations into Attempto Controlled English (ACE). One experienced pediatrician, one physician and a knowledge engineer assessed that a suitably extended version of ACE can accurately and naturally represent the clinical concepts and the proposed actions of the guidelines. Currently, we are developing a systematic and replicable approach to authoring guideline recommendations in ACE.

Richard N. Shiffman, George Michel, Michael Krauthammer, Norbert E. Fuchs, Kaarel Kaljurand, Tobias Kuhn

What Are Controlled Natural Languages?

On Controlled Natural Languages: Properties and Prospects

Abstract

This collaborative report highlights the properties and prospects of Controlled Natural Languages (CNLs). The report poses a range of questions concerning the goals of the CNL, the design, the linguistic aspects, the relationships and evaluation of CNLs, and the application tools. In posing the questions, the report attempts to structure the field of CNLs and to encourage further systematic discussion by researchers and developers.

Adam Wyner, Krasimir Angelov, Guntis Barzdins, Danica Damljanovic, Brian Davis, Norbert Fuchs, Stefan Hoefler, Ken Jones, Kaarel Kaljurand, Tobias Kuhn, Martin Luts, Jonathan Pool, Mike Rosner, Rolf Schwitter, John Sowa

Backmatter

Titel: Controlled Natural Language
herausgegeben von: Norbert E. Fuchs
Verlag: Springer Berlin Heidelberg
Electronic ISBN: 978-3-642-14418-9
Print ISBN: 978-3-642-14417-2
DOI: https://doi.org/10.1007/978-3-642-14418-9