BibSonomyburst/user/gromgullBibSonomy RSS feed for /user/gromgull2012-02-15T11:41:40+01:00http://www.bibsonomy.org/bibtex/21b6ec91f0fcab665d9dde00579bfdd52/gromgullgromgull2012-01-24T10:22:46+01:00feature-vector hashing indexing machine-learning <span class="authorEditorList"><a href="/author/Forman">George Forman</a>, and <a href="/author/Kirshenbaum">Evan Kirshenbaum</a> </span><em>CIKM &#039;08: Proceeding of the 17th ACM conference on Information and knowledge management, </em><em>page 1221--1230. </em><em>New York, NY, USA, </em><em>ACM, </em>(<em>2008</em>)Tue Jan 24 10:22:46 CET 2012New York, NY, USACIKM '08: Proceeding of the 17th ACM conference on Information and knowledge management1221--1230Extremely fast text feature extraction for classification and indexing2008feature-vector hashing indexing machine-learning Most research in speeding up text mining involves algorithmic improvements to induction algorithms, and yet for many large scale applications, such as classifying or indexing large document repositories, the time spent extracting word features from texts can itself greatly exceed the initial training time. This paper describes a fast method for text feature extraction that folds together Unicode conversion, forced lowercasing, word boundary detection, and string hash computation. We show empirically that our integer hash features result in classifiers with equivalent statistical performance to those built using string word features, but require far less computation and less memory.Extremely fast text feature extraction for classification and indexinghttp://www.bibsonomy.org/bibtex/2a0d72c90aa3348858a647e7603ad7323/gromgullgromgull2012-01-16T13:25:40+01:00dimensionality-reduction embedding machine-learning visualisation <span class="authorEditorList"><a href="/author/Hinton">Geoffrey Hinton</a>, and <a href="/author/Roweis">Sam Roweis</a> </span><em>Advances in neural information processing systems</em> (<em>2003</em>)Mon Jan 16 13:25:40 CET 2012Advances in neural information processing systems833--840Stochastic Neighbor Embedding152003dimensionality-reduction embedding machine-learning visualisation Stochastic Neighbor Embedding | Mendeleyhttp://www.bibsonomy.org/bibtex/23b3f7c23431f8a38572bf3d5133cc8ad/gromgullgromgull2012-01-16T10:50:59+01:00crf image-processing inference machinelearning <span class="authorEditorList"><a href="/author/Krahenbuhl">Philipp Krahenbuhl</a>, and <a href="/author/Koltun">Vladlen Koltun</a> </span>(<em>2011</em>)Mon Jan 16 10:50:59 CET 2012Neural Information Processing SystemsEfficient Inference in Fully Connected CRFs with Gaussian Edge Potentials2011crf image-processing inference machinelearning Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials | Mendeleyhttp://www.bibsonomy.org/bibtex/21339e16406730abfe5ad74ea49567253/gromgullgromgull2011-12-05T10:02:42+01:00disambiguation entity-matching graphical-models inference machine-learning map-reduce <span class="authorEditorList"><a href="/author/Singh">Sameer Singh</a>, <a href="/author/Subramanya">Amarnag Subramanya</a>, <a href="/author/Pereira">Fernando C. N. Pereira</a>, and <a href="/author/McCallum">Andrew McCallum</a> </span><em>ACL, </em><em>page 793-803. </em><em>The Association for Computer Linguistics, </em>(<em>2011</em>)Mon Dec 05 10:02:42 CET 2011ACLconf/acl/2011793-803Large-Scale Cross-Document Coreference Using Distributed Inference and Hierarchical Models.2011disambiguation entity-matching graphical-models inference machine-learning map-reduce http://www.bibsonomy.org/bibtex/25a9065e96237a69d95edebc03ccac92d/gromgullgromgull2011-11-23T17:03:54+01:00flickr hierarchy-learning machine-learning ontology ontology-learning <span class="authorEditorList"><a href="/author/Schmitz">Patrick Schmitz</a> </span><em>Collaborative Web Tagging Workshop at WWW 2006, </em><em>Edinburgh, Scotland, </em>(<em>May 2006</em>)Wed Nov 23 17:03:54 CET 2011Edinburgh, ScotlandCollaborative Web Tagging Workshop at WWW 2006mayInducing Ontology from Flickr Tags.2006flickr hierarchy-learning machine-learning ontology ontology-learning http://www.bibsonomy.org/bibtex/23b08719db9a27c9e8fa0d5d8f5c19a10/gromgullgromgull2011-11-23T16:31:28+01:00machine-learning nlp <span class="authorEditorList"><a href="/author/Lin">Dekang Lin</a> </span><em>Proceedings of the 17th international conference on Computational linguistics, </em><em>page 768--774. </em><em>Morristown, NJ, USA, </em><em>Association for Computational Linguistics, </em>(<em>1998</em>)Wed Nov 23 16:31:28 CET 2011Morristown, NJ, USAProceedings of the 17th international conference on Computational linguistics768--774Automatic retrieval and clustering of similar words1998machine-learning nlp Bootstrapping semantics from text is one of the greatest challenges in natural language learning. We first define a word similarity measure based on the distributional pattern of words. The similarity measure allows us to construct a thesaurus using a parsed corpus. We then present a new evaluation methodology for the automatically constructed thesaurus. The evaluation results show that the thesaurus is significantly closer to WordNet than Roget Thesaurus is.Automatic retrieval and clustering of similar wordshttp://www.bibsonomy.org/bibtex/250308d5168f519ce89a71fa67574ac25/gromgullgromgull2011-11-23T16:24:13+01:00distance-measure google machine-learning <span class="authorEditorList"><a href="/author/Cilibrasi">Rudi Cilibrasi</a>, and <a href="/author/Vitanyi">Paul M. B. Vitanyi</a> </span> (<em>March 2005</em>)<em>v2 . </em>Wed Nov 23 16:24:13 CET 201115 Marchv2cs.CL/0412098370-383Automatic Meaning Discovery Using Google2004distance-measure google machine-learning We have found a method to automatically extract the meaning of words and phrases from the world-wide-web using Google page counts. The approach is novel in its unrestricted problem domain, simplicity of implementation, and manifestly ontological underpinnings. The world-wide-web is the largest database on earth, and the latent semantic context information entered by millions of independent users averages out to provide automatic meaning of useful quality. We demonstrate positive correlations, evidencing an underlying semantic structure, in both numerical symbol notations and number-name words in a variety of natural languages and contexts. Next, we demonstrate the ability to distinguish between colours and numbers, and to distinguish between 17th century Dutch painters; the ability to understand electrical terms, religious terms, and emergency incidents; we conduct a massive experiment in understanding WordNet categories; and finally we demonstrate the ability to do a simple automatic English-Spanish translation.http://www.bibsonomy.org/bibtex/24f4881dbec5cb04b5889c9164097a10d/gromgullgromgull2011-11-18T12:12:17+01:00link-predication machinelearning readme record-linkage silk <span class="authorEditorList"><a href="/author/Isele">Robert Isele</a>, and <a href="/author/Bizer">Christian Bizer</a> </span><em>6th International Workshop on Ontology Matching. Bonn, </em>(<em>2011</em>)Fri Nov 18 12:12:17 CET 20116th International Workshop on Ontology Matching. BonnLearning Linkage Rules using Genetic Programming2011link-predication machinelearning readme record-linkage silk Silk - A Link Discovery Framework for the Web of Datahttp://www.bibsonomy.org/bibtex/27d0bd5b964c1fc33bd303c0ecb143d47/gromgullgromgull2011-10-14T15:24:09+02:00data machine-learning more-data natural-language-processing <span class="authorEditorList"><a href="/author/Banko">Michele Banko</a>, and <a href="/author/Brill">Eric Brill</a> </span><em>ACL &#039;01: Proceedings of the 39th Annual Meeting on Association for Computational Linguistics, </em><em>page 26--33. </em><em>Morristown, NJ, USA, </em><em>Association for Computational Linguistics, </em>(<em>2001</em>)Fri Oct 14 15:24:09 CEST 2011Morristown, NJ, USAACL '01: Proceedings of the 39th Annual Meeting on Association for Computational Linguistics26--33Scaling to very very large corpora for natural language disambiguation2001data machine-learning more-data natural-language-processing With a billion word corpus, your algorithm doesn't matter - and you can skip all your clever tricks. Also, active learning works better with huge data sets to pick interesting examples from. http://www.bibsonomy.org/bibtex/2b04843d465bcfbd35385410970c82473/gromgullgromgull2011-10-14T15:22:41+02:00data-mining rfid sensor-mining sensors <span class="authorEditorList"><a href="/author/Perkowitz">Mike Perkowitz</a>, <a href="/author/Philipose">Matthai Philipose</a>, <a href="/author/Fishkin">Kenneth Fishkin</a>, and <a href="/author/Patterson">Donald J. Patterson</a> </span><em>WWW &#039;04: Proceedings of the 13th international conference on World Wide Web, </em><em>page 573--582. </em><em>New York, NY, USA, </em><em>ACM, </em>(<em>2004</em>)Fri Oct 14 15:22:41 CEST 2011New York, NY, USAWWW '04: Proceedings of the 13th international conference on World Wide Web573--582Mining models of human activities from the web2004data-mining rfid sensor-mining sensors The ability to determine what day-to-day activity (such as cooking pasta, taking a pill, or watching a video) a person is performing is of interest in many application domains. A system that can do this requires models of the activities of interest, but model construction does not scale well: humans must specify low-level details, such as segmentation and feature selection of sensor data, and high-level structure, such as spatio-temporal relations between states of the model, for each and every activity. As a result, previous practical activity recognition systems have been content to model a tiny fraction of the thousands of human activities that are potentially useful to detect. In this paper, we present an approach to sensing and modeling activities that scales to a much larger class of activities than before. We show how a new class of sensors, based on Radio Frequency Identification (RFID) tags, can directly yield semantic terms that describe the state of the physical world. These sensors allow us to formulate activity models by translating labeled activities, such as 'cooking pasta', into probabilistic collections of object terms, such as 'pot'. Given this view of activity models as text translations, we show how to mine definitions of activities in an unsupervised manner from the web. We have used our technique to mine definitions for over 20,000 activities. We experimentally validate our approach using data gathered from actual human activity as well as simulated data.Mining models of human activities from the webhttp://www.bibsonomy.org/bibtex/2a67328ff553022857113ce7456a3d7cf/gromgullgromgull2011-08-12T10:19:48+02:00application machine-learning music neural-networks recurrent-neural-networks <span class="authorEditorList"><a href="/author/Eck">Douglas Eck</a>, and <a href="/author/Schmidhuber">Jürgen Schmidhuber</a> </span><em>NEURAL NETWORKS FOR SIGNAL PROCESSING XII, PROCEEDINGS OF THE 2002 IEEE WORKSHOP, </em><em>page 747--756. </em><em>IEEE, </em>(<em>2002</em>)Fri Aug 12 10:19:48 CEST 2011NEURAL NETWORKS FOR SIGNAL PROCESSING XII, PROCEEDINGS OF THE 2002 IEEE WORKSHOP747--756Finding temporal structure in music: Blues improvisation with LSTM recurrent networks2002application machine-learning music neural-networks recurrent-neural-networks Few types of signal streams are as ubiquitous as music. Here we consider the problem of extracting essential ingredients of music signals, such as well-defined global temporal structure in the form of nested periodicities (or meter). Can we construct an adaptive signal processing device that learns by example how to generate new instances of a given musical style? Because recurrent neural networks can in principle learn the temporal structure of a signal, they are good candidates for such a task. Unfortunately, music composed by standard recurrent neural networks (RNNs) often lacks global coherence. The reason for this failure seems to be that RNNs cannot keep track of temporally distant events that indicate global music structure. Long Short-Term Memory (LSTM) has succeeded in similar domains where other RNNs have failed, such as timing &amp; counting and learning of context sensitive languages. In the current study we show that LSTM is also a good mechanism for learning to compose music. We present experimental results showing that LSTM successfully learns a form of blues music and is able to compose novel (and we believe pleasing) melodies in that style. Remarkably, once the network has found the relevant structure it does not drift from it: LSTM is able to play the blues with good timing and proper structure as long as one is willing to listen.CiteSeerX — Finding temporal structure in music: Blues improvisation with LSTM recurrent networkshttp://www.bibsonomy.org/bibtex/2af0029f21446a26c04f2e4650ec1fbf1/gromgullgromgull2011-07-08T14:11:15+02:00machinelearning neural-networks optimisation recurrent-neural-networks <span class="authorEditorList"><a href="/author/Martens">James Martens</a> </span><em>ICML, </em><em>page 735-742. </em><em>Omnipress, </em>(<em>2010</em>)Fri Jul 08 14:11:15 CEST 2011ICMLconf/icml/2010735-742Deep learning via Hessian-free optimization.2010machinelearning neural-networks optimisation recurrent-neural-networks http://www.bibsonomy.org/bibtex/2c397781f6a6e59e824f620843712c3a0/gromgullgromgull2011-06-08T16:03:02+02:00book linked-open-data linkeddata <span class="authorEditorList"><a href="/author/Heath">Tom Heath</a>, and <a href="/author/Bizer">Christian Bizer</a> </span><em>Morgan \&amp; Claypool, </em><em>1st edition, </em>(<em>2011</em>)Wed Jun 08 16:03:02 CEST 20111stLinked Data: Evolving the Web into a Global Data Space2011book linked-open-data linkeddata The World Wide Web has enabled the creation of a global information space comprising linked documents. As the Web becomes ever more enmeshed with our daily lives, there is a growing desire for direct access to raw data not currently available on the Web or bound up in hypertext documents. Linked Data provides a publishing paradigm in which not only documents, but also data, can be a first class citizen of the Web, thereby enabling the extension of the Web with a global data space based on open standards - the Web of Data. In this Synthesis lecture we provide readers with a detailed technical introduction to Linked Data. We begin by outlining the basic principles of Linked Data, including coverage of relevant aspects of Web architecture. The remainder of the text is based around two main themes - the publication and consumption of Linked Data. Drawing on a practical Linked Data scenario, we provide guidance and best practices on: architectural approaches to publishing Linked Data; choosing URIs and vocabularies to identify and describe resources; deciding what data to return in a description of a resource on the Web; methods and frameworks for automated linking of data sets; and testing and debugging approaches for Linked Data deployments. We give an overview of existing Linked Data applications and then examine the architectures that are used to consume Linked Data from the Web, alongside existing tools and frameworks that enable these. Readers can expect to gain a rich technical understanding of Linked Data fundamentals, as the basis for application development, research or further study. Keywords: web technology, databases, linked data, web of data, semantic web, world wide web, dataspaces, data integration, data management, web engineering, resource description frameworkhttp://www.bibsonomy.org/bibtex/2d0889e5f137ca68dc1755769e63841dc/gromgullgromgull2011-06-01T15:55:02+02:00goodrelations ontology <span class="authorEditorList"><a href="/author/Hepp">Martin Hepp</a> </span><em>EKAW, </em><em>volume 5268 of Lecture Notes in Computer Science, </em><em>page 329-346. </em><em>Springer, </em>(<em>2008</em>)Wed Jun 01 15:55:02 CEST 2011EKAWconf/ekaw/2008329-346Lecture Notes in Computer ScienceGoodRelations: An Ontology for Describing Products and Services Offers on the Web.52682008goodrelations ontology dblphttp://www.bibsonomy.org/bibtex/2a42e758173e9de074287f4ff71362134/gromgullgromgull2011-05-24T12:14:52+02:00feature-vector hashing machine-learning <span class="authorEditorList"><a href="/author/Ganchev">Kuzman Ganchev</a>, and <a href="/author/Dredze">Mark Dredze</a> </span><em>Proceedings of the ACL-2008 Workshop on Mobile Language Processing, </em><em>Association for Computational Linguistics, </em>(<em>2008</em>)Tue May 24 12:14:52 CEST 2011Proceedings of the ACL-2008 Workshop on Mobile Language ProcessingSmall Statistical Models by Random Feature Mixing2008feature-vector hashing machine-learning Kuzman@Pennhttp://www.bibsonomy.org/bibtex/2b8c355086dd09a91893c7a643e9796f4/gromgullgromgull2011-05-17T21:14:34+02:00hashing label-propagation machine-learning <span class="authorEditorList"><a href="/author/Baluja">Shumeet Baluja</a>, <a href="/author/Ravichandran">Deepak Ravichandran</a>, and <a href="/author/Sivakumar">D. Sivakumar</a> </span><em>Proceeding of the International Conference on Knowledge Discovery and Information Retrieval KDIR 2009, </em><em>INSTICC, </em>(<em>Oct 6, 2009</em>)Tue May 17 21:14:34 CEST 2011Proceeding of the International Conference on Knowledge Discovery and Information Retrieval (KDIR 2009)octText Classification Through Time: Efficient Label Propagation in {Time-Based} Graphs2009hashing label-propagation machine-learning 6-8One of the fundamental assumptions for machine learning based text classification systems is that the underlying distribution from which the set of labeled-text is drawn is identical to the distribution from which the text-to-be-labeled is drawn. However, in live news aggregation sites, this assumption is rarely correct. Instead, the events and topics discussed in news stories dramatically change over time. Rather than ignoring this phenomenon, we attempt to explicitly model the transitions of news stories and classifications over time to label stories that may be acquired months after the initial examples are labeled. We test our system, based on efficiently propagating labels in time-based graphs, with recently published news stories collected over an eighty day period. Experiments presented in this paper include the use of training labels from each story within the first several days of gathering stories, to using a single story as a label.CiteULike: Text Classification Through Time: Efficient Label Propagation in Time-Based Graphshttp://www.bibsonomy.org/bibtex/236dece1b3288be384fe2bfeda7fb200d/gromgullgromgull2011-01-06T10:41:08+01:00machine-learning ontology-refactoring owl semantic-web toread <span class="authorEditorList"><a href="/author/Lehmann">Jens Lehmann</a>, and <a href="/author/Bühmann">Lorenz Bühmann</a> </span><em>Proceedings of the 9th International Semantic Web Conference ISWC2010, </em><em>Berlin / Heidelberg, </em><em>Springer, </em>(<em>2010</em>)Thu Jan 06 10:41:08 CET 2011Berlin / HeidelbergProceedings of the 9th International Semantic Web Conference (ISWC2010)Lecture Notes in Computer ScienceORE - A Tool for Repairing and Enriching Knowledge Bases2010machine-learning ontology-refactoring owl semantic-web toread http://www.bibsonomy.org/bibtex/25e4a1c252de27c4b01b137d8704631f8/gromgullgromgull2011-01-06T10:40:27+01:00ontology-refactoring semantic-web toread <span class="authorEditorList"><a href="/author/Rieß">Christoph Rieß</a>, <a href="/author/Heino">Norman Heino</a>, <a href="/author/Tramp">Sebastian Tramp</a>, and <a href="/author/Auer">Sören Auer</a> </span><em>Proceedings of the 9th International Semantic Web Conference ISWC2010, </em><em>Berlin / Heidelberg, </em><em>Springer, </em>(<em>2010</em>)Thu Jan 06 10:40:27 CET 2011Berlin / HeidelbergProceedings of the 9th International Semantic Web Conference (ISWC2010)Lecture Notes in Computer ScienceEvo{P}at -- {P}attern-{B}ased {E}volution and {R}efactoring of {RDF} {K}nowledge {B}ases2010ontology-refactoring semantic-web toread http://www.bibsonomy.org/bibtex/2c1cf4c2b50e8d9ae4b88714d6df7770f/gromgullgromgull2010-12-20T15:04:27+01:00dbtropes linked-open-data tvtropes <span class="authorEditorList"><a href="/author/Grimnes">Malte Kiesel; Gunnar Aastrand Grimnes</a> </span><em>EKAW 2010 Demo &amp; Poster Abstracts. International Conference on Knowledge Engineering and Knowledge Management EKAW-10, 17th International Conference on Knowledge Engineering and Knowledge Management, October 11-15, Lisbon, Portugal, </em><em>-, </em>(<em>October 2010</em>)<em>Best Poster . </em>Mon Dec 20 15:04:27 CET 2010EKAW 2010 Demo & Poster Abstracts. International Conference on Knowledge Engineering and Knowledge Management (EKAW-10), 17th International Conference on Knowledge Engineering and Knowledge Management, October 11-15, Lisbon, Portugal10Best PosterDBTropes---a linked data wrapper approach incorporating community feedback2010dbtropes linked-open-data tvtropes A common approach for serving Linked Data is to modify existing services to translate and export the underlying data as RDF. However, for many existing data sources on the web such an approach is not feasible: large installations might not be suitable for the changes necessary, programmers possibly are not able to adapt the software, or the data might not be suited for direct translation to RDF. DBTropes.org is a wrapper to TV Tropes, a wiki describing works of fiction by associating features---known as "Tropes". DBTropes is an independent service only using public data available via HTTP and translating it to RDF. Since the TV Tropes wiki does not provide structured data, the extracted data is noisy, and the interpretation of the data is sometimes ambiguous. DBTropes features a user interface that allows correcting and amending the data extracted from TV Tropes. This allows the extracted data to stay in sync with the original wiki, while also allowing the linked-data community to fix extraction error. http://www.bibsonomy.org/bibtex/2a87ba99ae5f55dcf677d0fd10723e733/gromgullgromgull2010-11-23T10:25:05+01:00feature-vector machine-learning semantic-web spreading-activation <span class="authorEditorList"><a href="/author/Teufl">Peter Teufl</a>, and <a href="/author/Lackner">Günther Lackner</a> </span><em>10th International Conference on Knowledge Management and Knowledge Technologies 1–3 September 2010, Messe Congress Graz, Austria, </em><em>page 18 - 18. </em>(<em>2010</em>)Tue Nov 23 10:25:05 CET 201010th International Conference on Knowledge Management and Knowledge Technologies 1–3 September 2010, Messe Congress Graz, Austria18 - 18Journal of Computer ScienceRDF Data Analysis with Activation Patterns2010feature-vector machine-learning semantic-web spreading-activation IAIK - TU Graz :: BibTeX