<html><head><meta http-equiv="Content-Type" content="text/html charset=windows-1252"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;">We are pleased to announce the electronic publication of Language Resources and Evaluation, Volume 48, Issue 3, available on <a href="http://alerts.springer.com/re?l=D0In5s91fI6gwaaktIg">SpringerLink</a><div><br></div><div><div style="text-align: left;"><b> Language Resources and Evaluation</b></div><div style="text-align: left;"> Volume 48, Issue 3</div><div style="text-align: left;"><br></div><div style="text-align: left;"><b>Table of Contents</b></div><div><br></div><div><i>The Linguistic Annotation Framework: a standard for annotation interchange and</i></div><div><i>merging</i> </div><div>Nancy Ide & Keith Suderman</div><div><br></div><div>This paper overviews the International Standards Organization–Linguistic</div><div>Annotation Framework (ISO–LAF) developed in ISO TC37 SC4. We describe the XML</div><div>serialization of ISO–LAF, the Graph Annotation Format (GrAF) and discuss the</div><div>rationale behind the various decisions that were made in determining the</div><div>standard. We describe the structure of the GrAF headers in detail and provide</div><div>multiple examples of GrAF representation for text and multi-media. Finally, we</div><div>discuss the next steps for standardization of interchange formats for linguistic</div><div>annotations.</div><div><br></div><div><i>Automatic dialogue act recognition with syntactic features </i></div><div>Pavel Král & Christophe Cerisara</div><div><br></div><div>This work studies the usefulness of syntactic information in the context of</div><div>automatic dialogue act recognition in Czech. Several pieces of evidence are</div><div>presented in this work that support our claim that syntax might bring valuable</div><div>information for dialogue act recognition. In particular, a parallel is drawn</div><div>with the related domain of automatic punctuation generation and a set of</div><div>syntactic features derived from a deep parse tree is further proposed and</div><div>successfully used in a Czech dialogue act recognition system based on</div><div>conditional random fields. We finally discuss the possible reasons why so few</div><div>works have exploited this type of information before and propose future research</div><div>directions to further progress in this area.</div><div><br></div><div><i>A dependency annotation scheme for Bangla treebank</i> </div><div>Sanjay Chatterji, Tanaya Mukherjee Sarkar, Pragati Dhang, Samhita Deb, Sudeshna Sarkar, Jayshree</div><div>Chakraborty & Anupam Basu</div><div><br></div><div>Dependency grammar is considered appropriate for many Indian languages. In this</div><div>paper, we present a study of the dependency relations in Bangla language. We</div><div>have categorized these relations in three different levels, namely intrachunk</div><div>relations, interchunk relations and interclause relations. Each of these levels</div><div>is further categorized and an annotation scheme has been developed. Both</div><div>syntactic and semantic features have been taken into consideration for</div><div>describing the relations. In our scheme, there are 63 such syntactico–semantic</div><div>relations. We have verified the scheme by tagging a corpus of 4167 Bangla</div><div>sentences to create a treebank (KGPBenTreebank).</div><div><br></div><div><b>Project Notes</b></div><div><br></div><div><i>Lexicon+TX: rapid construction of a multilingual lexicon with under-resourced</i></div><div><i>languages </i></div><div>Lian Tze Lim, Lay-Ki Soon, Tek Yong Lim, Enya Kong Tang & Bali</div><div>Ranaivo-Malançon</div><div><br></div><div>Most efforts at automatically creating multilingual lexicons require input</div><div>lexical resources with rich content (e.g. semantic networks, domain codes,</div><div>semantic categories) or large corpora. Such material is often unavailable and</div><div>difficult to construct for under-resourced languages. In some cases,</div><div>particularly for some ethnic languages, even unannotated corpora are still in</div><div>the process of collection. We show how multilingual lexicons with</div><div>under-resourced languages can be constructed using simple bilingual translation</div><div>lists, which are more readily available. The prototype multilingual lexicon</div><div>developed comprise six member languages: English, Malay, Chinese, French, Thai</div><div>and Iban, the last of which is an under-resourced language in Borneo. Quick</div><div>evaluations showed that 91.2 % of 500 random multilingual entries in the</div><div>generated lexicon require minimal or no human correction.</div><div><br></div><div><i>Building the essential resources for Finnish: the Turku Dependency Treebank</i></div><div>Katri Haverinen, Jenna Nyblom, Timo Viljanen, Veronika Laippala, Samuel</div><div>Kohonen, Anna Missilä, Stina Ojala, Tapio Salakoski & Filip Ginter</div><div><br></div><div>In this paper, we present the final version of a publicly available treebank of</div><div>Finnish, the Turku Dependency Treebank. The treebank contains 204,399 tokens</div><div>(15,126 sentences) from 10 different text sources and has been manually</div><div>annotated in a Finnish-specific version of the well-known Stanford Dependency</div><div>scheme. The morphological analyses of the treebank have been assigned using a</div><div>novel machine learning method to disambiguate readings given by an existing</div><div>tool. As the second main contribution, we present the first open source Finnish</div><div>dependency parser, trained on the newly introduced treebank. The parser achieves</div><div>a labeled attachment score of 81 %. The treebank data as well as the parsing</div><div>pipeline are available under an open license at <a href="http://bionlp.utu.fi/">http://bionlp.utu.fi/</a>.</div><div><br></div><div><b>Book Review </b></div><div><b><br></b></div><div>I. Mani and J. Pustejovsky: <i>Interpreting motion: grounded representations for spatial language </i></div><div>Giovanna Marotta</div><div apple-content-edited="true"><br class="Apple-interchange-newline">
</div>
<br></div></body></html>