04.11.2012 Views

Dissertation Proposal: Design and ... - IMS - Universität Stuttgart

Dissertation Proposal: Design and ... - IMS - Universität Stuttgart

Dissertation Proposal: Design and ... - IMS - Universität Stuttgart

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

<strong>Dissertation</strong> <strong>Proposal</strong>:<br />

<strong>Design</strong> <strong>and</strong> Implementation of a Data Model for a<br />

Multifunctional Electronic Dictionary<br />

Dennis Spohr<br />

Dennis.Spohr@web.de<br />

May 2006<br />

1 Introduction <strong>and</strong> Motivation<br />

The structure of electronic dictionaries is a central topic in lexicographic research, <strong>and</strong> a<br />

variety of approaches have been pursued <strong>and</strong> implemented over the past decades. There<br />

is, however, an apparent gap between dictionaries for human use <strong>and</strong> those designed<br />

for applications of Natural Language Processing (NLP). On the one h<strong>and</strong>, there exist<br />

dictionaries such as the Valency Dictionary of English (Vde; Herbst et al., 2004) which<br />

are targeted solely at human users. The most obvious characteristic such resources<br />

have is that they contain large amounts of unformatted text – in the sense that this<br />

text does not contain any linguistic markup or meta-information – <strong>and</strong> very few (if<br />

any) links between lexical entries. The consequence is that considerable effort has to<br />

be made in order to make the contained information usable for NLP applications (see<br />

e.g. Spohr, 2004; Asmussen <strong>and</strong> Ørsnes, 2005). On the other h<strong>and</strong>, there are resources<br />

such as subcategorisation lexica for formal grammars (e.g. LFG, Lexical Functional<br />

Grammar) or the German morphology dictionary ImsLex (Fitschen, 2004), which have<br />

been designed specifically for one type of NLP application or cover exactly one sort of<br />

linguistic information, i.e. they are very specific.<br />

The main research aim of the proposed thesis is to fill exactly this gap between<br />

dictionaries for human users <strong>and</strong> those designed for specific NLP applications, <strong>and</strong> hence<br />

to make a contribution to the notion of a multifunctional electronic dictionary – as<br />

suggested by Heid <strong>and</strong> Gouws (to appear) in the context of collocations – which is to<br />

serve both human users <strong>and</strong> different NLP applications.<br />

While the primary prerequisite that an electronic dictionary has to meet with respect<br />

to NLP use is to employ a powerful, expressive <strong>and</strong> (ideally) st<strong>and</strong>ardised formalism, one<br />

of the main requirements for a human use dictionary is that it should go beyond merely<br />

being an electronic version of a printed dictionary. To make this possible, it is necessary<br />

to capture the highly relational nature of lexical data contained in a dictionary, i.e. to<br />

establish connections between the entities that are described. These connections are<br />

however not limited to lexical items, but may also involve links between the descriptions<br />

of such items, e.g. between syntatic <strong>and</strong> semantic descriptions. For those connections to<br />

be fully explored, it is necessary to provide fast <strong>and</strong> easy retrieval access to the data in<br />

the model without limiting the ways in which explicit (or even implicit) information can<br />

be queried. In addition to the st<strong>and</strong>ard way in which lexical information contained in<br />

an electronic dictionary is accessed, i.e. via the lemma, this issue involves the possibility<br />

1


of non-st<strong>and</strong>ard access to lexical information, e.g. via combinations of morphological,<br />

syntactic <strong>and</strong>/or semantic criteria (“retrieve the nominalisations of those verbs which<br />

are synonyms of . . . ”). With these requirements at the centre of investigation, one could<br />

move away from viewing an electronic dictionary primarily as a collection of lemmata<br />

combined with illustrative material (which is what many dictionaries directed towards<br />

human users are) <strong>and</strong> towards using it as a network-like instrument of linguistic research<br />

which is capable of not only representing lexical knowledge but also serving as a tool to<br />

identify new connections <strong>and</strong> relations between lexical items.<br />

To cover the multilingual dimension of lexical knowledge <strong>and</strong> of many of its applications,<br />

the research questions that have been addressed so far should be investigated with<br />

a view on multilinguality, i.e. considering how the described phenomena behave across<br />

various languages. The proposed dissertation would not, however, aim at providing a<br />

new formalism of semantic description for multilingual NLP, but would rather involve<br />

an analysis of existing formalisms <strong>and</strong> implementations (see below).<br />

2 Research Questions<br />

The research questions mentioned in the previous section are repeated below, grouped<br />

according to the type of requirement they impose on the dictionary model:<br />

• Linguistic requirements:<br />

– capture the relational nature of lexical data <strong>and</strong> the interaction between different<br />

levels of description<br />

– generic modelling of linguistic phenomena, i.e. a modelling not bound to a<br />

given single type of application<br />

• User/Usage requirements:<br />

– usability for both NLP applications <strong>and</strong> human users<br />

– non-st<strong>and</strong>ard search <strong>and</strong> retrieval access<br />

• Formal requirements:<br />

– powerful <strong>and</strong> expressive st<strong>and</strong>ard formalism, e.g. OWL DL 1 (W3C recommendation;<br />

Bechhofer et al., 2004)<br />

– compatibility with current best practice <strong>and</strong> st<strong>and</strong>ards<br />

– querying should be (largely) independent of the particular model structure<br />

The aim of this research is to design <strong>and</strong> implement a dictionary model which is<br />

to serve both NLP applications <strong>and</strong> human users, thus contributing to the notion of a<br />

multifunctional electronic dictionary. In addition to this, the model should be populated<br />

with data – ideally extracted from several existing resources – <strong>and</strong> evaluated according<br />

to its capability of achieving this goal, e.g. by implementing exemplary interfaces to<br />

(probably no more than) two NLP applications (e.g. parsing) <strong>and</strong> by designing views<br />

on the data with respect to the interests of two different groups of users (e.g. learners<br />

or translation).<br />

1 OWL DL is the Description Logic sublanguage of the Web Ontology Language OWL, which combines<br />

the expressiveness of OWL with the advantageous computational properties of Description Logics,<br />

e.g. non-monotonicity <strong>and</strong> decidability (see Baader et al. (2003) for a detailed account of the properties<br />

of different Description Logic languages <strong>and</strong> their use for various kinds of applications).<br />

2


3 Background <strong>and</strong> Related Work<br />

The Dictionary Writing System iLex (Erl<strong>and</strong>sen (2004); marketed by Erl<strong>and</strong>sen Media<br />

Publishing) is a commercial product offering highly efficient query access to very large<br />

XML (Schema) dictionaries. Some aspects of the query mechanism employed in iLex<br />

could be useful for the tasks mentioned in the previous section.<br />

Protégé (Knublauch et al., 2004) <strong>and</strong> the Sesame framework (Broekstra et al., 2002)<br />

provide attractive solutions for creating, storing <strong>and</strong> querying Description Logic ontologies<br />

by transforming them into a graph representation using RDF triples. In this<br />

formalisation, stating queries on a model corresponds to specifying paths through this<br />

graph. The use of these tools will be of considerable relevance to the implementation of<br />

the dictionary model.<br />

Mile (Calzolari et al., 2003) <strong>and</strong> the currently evolving Lexical Markup Framework<br />

(Francopoulo et al., 2006) aim at describing st<strong>and</strong>ards for the definition of lexical<br />

resources for NLP, in particular electronic dictionaries. However, they do not account<br />

for the highly relational (i.e. network-like) character of lexical data. As was already<br />

mentioned above, this relational aspect would be a central topic of investigation.<br />

The recently introduced Lexical Systems (Polguère, to appear) represent a kind of<br />

network-like (graph) model which can be generated from existing databases. A current<br />

implementation has been extracted from the DiCo 2 database, mainly describing lexemes<br />

according to lexical functions of Mel’čuk’s Meaning⇔Text-Theory (see e.g. Mel’čuk,<br />

1970). A feature which seems to be absent from this approach, however, is the distinction<br />

between object language <strong>and</strong> metalanguage vocabulary, which I believe is crucial<br />

e.g. with respect to consistency checking.<br />

Existing knowledge bases, such as WordNet (Fellbaum, 1998), Cyc (Lenat, 1995)<br />

or Sumo (Niles <strong>and</strong> Pease, 2001), contain taxonomic information about words <strong>and</strong> the<br />

concepts they denote. The classifications provided by such resources might be relevant<br />

for the description of semantic relations (e.g. synonymy) between lexical items. Attempts<br />

at combining subcategorisation information, WordNet, Cyc <strong>and</strong> VerbNet (Kipper<br />

et al., 2000) have been described in Crouch <strong>and</strong> King (2005).<br />

The Papillon multilingual dictionary architecture (Sérasset <strong>and</strong> Mangeot-Lerebours,<br />

2001) is based on the DiCo formalisation <strong>and</strong> is thus a model providing various kinds of<br />

relations between lexemes. Its account of multi-word expressions is however very limited<br />

(very often as part of the entry of one of the components), as is query access to the data<br />

contained in the dictionary.<br />

The SIMuLLDA Lexical Database Application (Janssen, 2002) proposes a multilingual<br />

architecture for dictionaries using Formal Concept Analysis (FCA) as basic formalism<br />

for the definition of a Structured Interlingua. However, the focus in this work is<br />

mainly on semantic description <strong>and</strong> the acquisition of concepts from existing dictionaries<br />

(using FCA) <strong>and</strong> less emphasis is put on the interfaces between different linguistic<br />

descriptions.<br />

4 Own Work in the Field<br />

In my Diploma Thesis (Spohr, 2005), I have designed a Description Logic data model for<br />

a monolingual collocation dictionary, based on the Web Ontology Language OWL DL.<br />

The main focus was on providing an adequate description of collocations <strong>and</strong> related<br />

2 http://olst.ling.umontreal.ca/dicouebe/<br />

3


phenomena, such as the modelling of collocation types <strong>and</strong> of morphosyntactic preferences<br />

of collocations. However, this focus on collocations has often been at the cost<br />

of a detailed account of other linguistic phenomena of lexicographic relevance. These<br />

include, for instance, a detailed description of single-word lexemes <strong>and</strong> their relations to<br />

other entities in the dictionary, e.g. morphological <strong>and</strong> lexical semantic relations.<br />

Nonetheless, the work presented in the Diploma Thesis has been able to show that<br />

OWL DL offers a number of attractive solutions for consistent modelling of lexical data,<br />

which has been further emphasised in Spohr <strong>and</strong> Heid (2006) with respect to multilingual<br />

data. Thus, I would use several aspects of the data model developed there as a starting<br />

point <strong>and</strong> further elaborate on the representation of the relations between phenomena<br />

from different levels of linguistic description, such as the connection between verbs <strong>and</strong><br />

their nominalisations. Further emphasis would be on the integration into the framework<br />

of other NLP applications (e.g. specification of interfaces) <strong>and</strong> on fast <strong>and</strong> easy retrieval<br />

access to the information in the model, in order to advance towards a thorough networklike<br />

data model for a multifunctional electronic dictionary which is to serve both NLP<br />

<strong>and</strong> human users.<br />

In addition to this, I am currently involved in research activities with a group of<br />

researchers in the Salsa project at Saarl<strong>and</strong> University (Erk et al., 2003). One of the<br />

main objectives is the modelling of a lexicon which provides adequate representations of<br />

Salsa’s corpus data <strong>and</strong> their annotations as well as their relation to the classification<br />

defined in the FrameNet project (Baker et al., 1998). The modelling of this lexicon<br />

involves (frame-)semantic <strong>and</strong> syntactic descriptions as well as – to some extent – annotations<br />

<strong>and</strong> classifications of certain kinds of metaphorical uses in corpus sentences.<br />

The knowledge gained during these activities will certainly be extremely relevant to the<br />

proposed task.<br />

5 Structure of the <strong>Dissertation</strong><br />

5.1 Specific objectives <strong>and</strong> approaches<br />

The specific steps towards achieving the goals mentioned in section 2 are listed below,<br />

subdivided into several objectives.<br />

Analysis of current research, especially in the fields of general linguistics, lexicography<br />

<strong>and</strong> data modelling in order to get a clear picture of the requirements<br />

<strong>and</strong> possibilities of lexical resource modelling, e.g. with respect to st<strong>and</strong>ards (see<br />

above).<br />

<strong>Design</strong> of a data model for a multifunctional electronic dictionary, primarily based<br />

on the preceding analysis step. Emphasis will be on the modelling of the interrelation<br />

of different levels of linguistic description (e.g. valence information of<br />

morphologically related lexemes), focussing mainly on monolingual descriptions,<br />

however, with a view to the requirements of multilingual lexical data.<br />

Implementation of an application for usage <strong>and</strong> persistent storage of this data model<br />

(e.g. database), which would most probably be based on the Sesame framework<br />

(cf. Broekstra et al., 2002). Further effort would go into the design of interfaces<br />

for connecting the data model to the frameworks of other NLP applications.<br />

Population of the model in order to be able to evaluate the model to some extent<br />

<strong>and</strong> to provide a solid basis for the following steps. This objective could probably<br />

4


draw on data from Vde (for valence descriptions; Herbst et al., 2004), FrameNet<br />

(for semantic features <strong>and</strong> especially valence patterns; Baker et al., 1998), Salsa<br />

(similar to FrameNet, but for German; Erk et al., 2003), GermaNet (Hamp<br />

<strong>and</strong> Feldweg, 1997) <strong>and</strong> (Euro)WordNet (for taxonomic relations among concepts;<br />

Fellbaum, 1998), ImsLex (for morphological information; Fitschen, 2004),<br />

<strong>and</strong> possibly further corpus data available at <strong>IMS</strong> (cf. Ritz, 2005). The information<br />

extracted from these resources would have to be merged in order to be fully<br />

exploited. One of the ways to do this has been described by Crouch <strong>and</strong> King<br />

in their approach to building a Unified Lexicon by merging WordNet, Cyc <strong>and</strong><br />

further external resources (cf. Crouch <strong>and</strong> King, 2005). Integrating data from<br />

different resources with focus on (largely) different aspects of linguistic analysis<br />

would be ideal for the evaluation of the model’s ability to capture the interrelation<br />

of different levels of linguistic description.<br />

Enhancement of the querying facility for easier access to the information in the<br />

model. The intention would be to be as generic as possible, i.e. independent of<br />

a particular model structure. Thus, a user or program should be able to query<br />

the underlying model even with minimal prerequisite knowledge of its structure.<br />

Instead of designing a single static query language for this purpose, a possible<br />

approach could be to develop some kind of semi-automatic query language generator<br />

which allows for the definition of path aliases or short-cuts to information<br />

in the model. This objective would require an analysis of graph traversal <strong>and</strong><br />

concordance computation algorithms (see e.g. Bohnet (to appear)).<br />

<strong>Design</strong> of views on the data in the model as well as the presentation thereof, depending<br />

on the needs <strong>and</strong> interests of a particular application (see above) or group of<br />

users. This would very likely be limited to the specification of no more than two<br />

of each kind, with comments on the general procedure.<br />

5.2 Possible schedule<br />

The table below proposes a schedule for the dissertation. The actual procedure might<br />

deviate from this planning, as some of the objectives might require simultaneous processing.<br />

However, the rough order should correspond to the one specified below.<br />

Year<br />

Objective 1 st 2 nd 3 rd<br />

<strong>Design</strong> of the dictionary data model × (×)<br />

Implementation <strong>and</strong> model population (×) ×<br />

Enhancement of the querying facility × (×)<br />

<strong>Design</strong> of views on the data ×<br />

Writing (×) (×) ×<br />

5


References<br />

Asmussen, Jørg <strong>and</strong> Bjarne Ørsnes, 2005. Valency information for dictionaries <strong>and</strong> NLP<br />

lexicons: Adapting valency frames from The Danish Dictionary to an LFG lexicon.<br />

In Ferenc Kiefer <strong>and</strong> Júlia Pajzs (eds.), Proceedings of the 8th Conference on Computational<br />

Lexicography (COMPLEX 2005) – Papers in Computational Lexicography.<br />

Budapest, Hungary.<br />

Baader, Franz, Diego Calvanese, Deborah L. McGuinness, Daniele Nardi, <strong>and</strong> Peter F.<br />

Patel-Schneider, 2003. The Description Logic H<strong>and</strong>book: Theory, Implementation <strong>and</strong><br />

Applications. Cambridge, UK: Cambridge University Press.<br />

Baker, Collin F., Charles Fillmore, <strong>and</strong> John B. Lowe, 1998. The Berkeley FrameNet<br />

project. In Proceedings of COLING-ACL. Montreal, Canada.<br />

Bechhofer, Sean, Frank van Harmelen, Jim Hendler, Ian Horrocks, Deborah L. McGuinness,<br />

Peter F. Patel-Schneider, <strong>and</strong> Lynn Andrea Stein, 2004. OWL Web Ontology<br />

Language Reference. Technical report. W3C. http://www.w3.org/TR/owl-ref/. Last<br />

accessed: 15.05.2006.<br />

Bohnet, Bernd, to appear. Textgenerierung durch Transduktion linguistischer Strukturen.<br />

Doctoral <strong>Dissertation</strong>. Institut für Intelligente Systeme, <strong>Universität</strong> <strong>Stuttgart</strong>,<br />

Germany.<br />

Broekstra, Jeen, Arjohn Kampman, <strong>and</strong> Frank van Hermelen, 2002. Sesame: A Generic<br />

Architecture for Storing <strong>and</strong> Querying RDF <strong>and</strong> RDF Schema. In Ian Horrocks <strong>and</strong><br />

James Hendler (eds.), Proceedings of the First International Semantic Web Conference<br />

(ISWC 2002). Sardinia, Italy.<br />

Calzolari, Nicoletta, Francesca Bertagna, Aless<strong>and</strong>ro Lenci, <strong>and</strong> Monica Monachini,<br />

2003. St<strong>and</strong>ards <strong>and</strong> Best Practice for Multilingual Computational Lexicons & MILE<br />

(the Multilingual ISLE Lexical Entry). ISLE Computational Lexicon Working Group,<br />

Università di Pisa, Italy.<br />

Crouch, Dick <strong>and</strong> Tracy Holloway King, 2005. Unifying Lexical Resources. In Katrin<br />

Erk, Alissa Melinger, <strong>and</strong> Sabine Schulte im Walde (eds.), Proceedings of the Interdisciplinary<br />

Workshop on the Identification <strong>and</strong> Representation of Verb Features <strong>and</strong><br />

Verb Classes. Saarbrücken, Germany.<br />

Erk, Katrin, Andrea Kowalski, <strong>and</strong> Manfred Pinkal, 2003. A corpus resource for lexical<br />

semantics. In Harry Bunt, Ielka van der Sluis, <strong>and</strong> Roser Morante (eds.), Proceedings<br />

of the fifth International Workshop on Computational Semantics (IWCS-5). Tilburg,<br />

The Netherl<strong>and</strong>s.<br />

Erl<strong>and</strong>sen, Jens, 2004. iLex – new DWS. In Pavel Smrz, Philippe Climent, Steve Crowdy,<br />

Frantisek Cermak, Laura Elliott, Adam Kilgarriff, <strong>and</strong> Gilles-Maurice de Schryver<br />

(eds.), Proceedings of the 3rd International Workshop on Dictionary Writing Systems<br />

(DWS-2004). Brno, Czech Republic.<br />

Fellbaum, Christiane, 1998. WordNet: An Electronic Lexical Database. Cambridge,<br />

Massachusetts: MIT Press.<br />

6


Fitschen, Arne, 2004. Ein Computerlinguistisches Lexikon als komplexes System. Doctoral<br />

<strong>Dissertation</strong>. Arbeitspapiere des Instituts für Maschinelle Sprachverarbeitung<br />

(A<strong>IMS</strong>), Volume 10(3). Institut für Maschinelle Sprachverarbeitung, <strong>Universität</strong><br />

<strong>Stuttgart</strong>, Germany.<br />

Francopoulo, Gil, Monte George, Nicoletta Calzolari, Monica Monachini, Nuria Bel,<br />

M<strong>and</strong>y Pet, <strong>and</strong> Claudia Soria, 2006. Lexical Markup Framework (LMF). In Nicoletta<br />

Calzolari, Khalid Choukri, Aldo Gangemi, Bente Maegaard, Joseph Mariani, Jan<br />

Odijk, <strong>and</strong> Daniel Tapias (eds.), Proceedings of the 5th International Conference on<br />

Lexical Resources <strong>and</strong> Evaluation (LREC-2006). Genoa, Italy.<br />

Hamp, Birgit <strong>and</strong> Helmut Feldweg, 1997. GermaNet – a Lexical-Semantic Net for<br />

German. In Piek Vossen, Geert Adriaens, Nicoletta Calzolari, Antonio Sanfilippo,<br />

<strong>and</strong> Yorick Wilks (eds.), Proceedings of the ACL Workshop on Automatic Information<br />

Extraction <strong>and</strong> Building of Lexical Semantic Resources for NLP Applications. Madrid,<br />

Spain.<br />

Heid, Ulrich <strong>and</strong> Rufus H. Gouws, to appear. A model for a multifunctional electronic<br />

dictionary of collocations. To appear in Proceedings of the 12th EURALEX International<br />

Congress (EURALEX-2006). Turin, Italy.<br />

Herbst, Thomas, David Heath, Ian Roe, <strong>and</strong> Dieter Götz, 2004. A Valency Dictionary<br />

of English. Berlin/New York: de Gruyter.<br />

Janssen, Maarten, 2002. SIMuLLDA: a Multilingual Lexical Database Application using<br />

a Structured Interlingua. PhD Thesis. Utrecht Institute of Linguistics OTS, Universiteit<br />

Utrecht, The Netherl<strong>and</strong>s.<br />

Kipper, Karin, Hoa Trang Dang, <strong>and</strong> Martha Palmer, 2000. Class-based Construction<br />

of a Verb Lexicon. In Harry Bunt, Ielka van der Sluis, <strong>and</strong> Roser Morante (eds.),<br />

Proceedings of the Seventeenth National Conference on Artificial Intelligence (AAAI-<br />

2000). Austin, TX, USA.<br />

Knublauch, Holger, Ray W. Fergerson, Natalya F. Noy, <strong>and</strong> Mark A. Musen, 2004.<br />

The Protégé OWL Plugin: An Open Development Environment for Semantic Web<br />

Applications. In Sheila A. McIlraith, Dimitris Plexousakis, <strong>and</strong> Frank van Harmelen<br />

(eds.), Proceedings of the Third International Semantic Web Conference - ISWC 2004 .<br />

Hiroshima, Japan.<br />

Lenat, Doug, 1995. Cyc: a large-scale investment in knowledge infrastructure. In CACM<br />

38, No. 11 .<br />

Mel’čuk, Igor A., 1970. Towards a functioning model of language. In Manfred Bierwisch<br />

<strong>and</strong> Karl Erich Heidolph (eds.), Progress in Linguistics. Den Haag, The Netherl<strong>and</strong>s:<br />

Mouton.<br />

Niles, Ian <strong>and</strong> Adam Pease, 2001. Towards a St<strong>and</strong>ard Upper Ontology. In Chris Welty<br />

<strong>and</strong> Barry Smith (eds.), Proceedings of the 2nd International Conference on Formal<br />

Ontology in Information Systems (FOIS-2001). Ogunquit, Maine.<br />

Polguère, Alain, to appear. Structural properties of Lexical Systems: Monolingual <strong>and</strong><br />

Multilingual Perspectives. Submitted to COLING/ACL Workshop on Multilingual<br />

Language Resources <strong>and</strong> Interoperability. Sydney, Australia.<br />

7


Ritz, Julia, 2005. Entwicklung eines Systems zur Extraktion von Kollokationen mittels<br />

morphosyntaktischer Features. Diploma Thesis. Institut für Maschinelle Sprachverarbeitung,<br />

<strong>Universität</strong> <strong>Stuttgart</strong>, Germany.<br />

Sérasset, Gilles <strong>and</strong> Mathieu Mangeot-Lerebours, 2001. Papillon lexical database<br />

project: Monolingual dictionaries & interlingual links. In Keh-Yih Su, Yuji Matsumoto,<br />

<strong>and</strong> Tien-Shun Yao (eds.), Proceedings of the Sixth Natural Language Processing<br />

Pacific Rim Symposium: NLPRS-2001 . Tokyo, Japan.<br />

Spohr, Dennis, 2004. Using A Valency Dictionary of English to enhance the Lexicon of an<br />

English LFG Grammar. Studienarbeit. Institut für Maschinelle Sprachverarbeitung,<br />

<strong>Universität</strong> <strong>Stuttgart</strong>, Germany.<br />

Spohr, Dennis, 2005. A Description Logic Approach to Modelling Collocations. Diploma<br />

Thesis. Institut für Maschinelle Sprachverarbeitung, <strong>Universität</strong> <strong>Stuttgart</strong>, Germany.<br />

Spohr, Dennis <strong>and</strong> Ulrich Heid, 2006. Modeling Monolingual <strong>and</strong> Bilingual Collocation<br />

Dictionaries in Description Logics. In Paul Rayson, Serge Sharoff, <strong>and</strong> Svenja<br />

Adolphs (eds.), Proceedings of the EACL Workshop on Multi-Word-Expressions in a<br />

Multilingual Context (EACL-2006). Trento, Italy.<br />

8

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!