Search (68 results, page 2 of 4)

0.00418391 = product of:
  0.058574736 = sum of:
    0.058574736 = weight(_text_:elektronische in 1714) [ClassicSimilarity], result of:
      0.058574736 = score(doc=1714,freq=2.0), product of:
        0.14013545 = queryWeight, product of:
          4.728978 = idf(docFreq=1061, maxDocs=44218)
          0.029633347 = queryNorm
        0.41798657 = fieldWeight in 1714, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          4.728978 = idf(docFreq=1061, maxDocs=44218)
          0.0625 = fieldNorm(doc=1714)
  0.071428575 = coord(1/14)

Theme

Elektronische Dokumente

0.003790876 = product of:
  0.02653613 = sum of:
    0.00856136 = weight(_text_:information in 1867) [ClassicSimilarity], result of:
      0.00856136 = score(doc=1867,freq=4.0), product of:
        0.052020688 = queryWeight, product of:
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.029633347 = queryNorm
        0.16457605 = fieldWeight in 1867, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.046875 = fieldNorm(doc=1867)
    0.01797477 = weight(_text_:retrieval in 1867) [ClassicSimilarity], result of:
      0.01797477 = score(doc=1867,freq=2.0), product of:
        0.08963835 = queryWeight, product of:
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.029633347 = queryNorm
        0.20052543 = fieldWeight in 1867, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.046875 = fieldNorm(doc=1867)
  0.14285715 = coord(2/14)

Abstract

Purpose - The purpose of this paper is to survey the treatment of relationships, relationship expressions and the ways in which they manifest themselves in image descriptions. Design/methodology/approach - The term "relationship" is construed in the broadest possible way to include spatial relationships ("to the right of"), temporal ("in 1936," "at noon"), meronymic ("part of"), and attributive ("has color," "has dimension"). The intentions of these vaguely delimited categories with image information, image creation, and description in libraries and archives is complex and in need of explanation. Findings - The review brings into question many generally held beliefs about the relationship problem such as the belief that the semantics of relationships are somehow embedded in the relationship term itself and that image search and retrieval solutions can be found through refinement of word-matching systems. Originality/value - This review has no hope of systematically examining all evidence in all disciplines pertaining to this topic. It instead focusses on a general description of a theoretical treatment in Library and Information Science.

0.0037468998 = product of:
  0.026228298 = sum of:
    0.0050448296 = weight(_text_:information in 3667) [ClassicSimilarity], result of:
      0.0050448296 = score(doc=3667,freq=2.0), product of:
        0.052020688 = queryWeight, product of:
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.029633347 = queryNorm
        0.09697737 = fieldWeight in 3667, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.0390625 = fieldNorm(doc=3667)
    0.021183468 = weight(_text_:retrieval in 3667) [ClassicSimilarity], result of:
      0.021183468 = score(doc=3667,freq=4.0), product of:
        0.08963835 = queryWeight, product of:
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.029633347 = queryNorm
        0.23632148 = fieldWeight in 3667, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.0390625 = fieldNorm(doc=3667)
  0.14285715 = coord(2/14)

Abstract

Descriptive metadata play a key role in finding relevant search results in large amounts of unstructured data. However, current scientific audiovisual media are provided with little metadata, which makes them hard to find, let alone individual sequences. In this paper, the TIB / AV-Portal is presented as a use case where methods concerning the automatic generation of metadata, a semantic search and cross-lingual retrieval (German/English) have already been applied. These methods result in a better discoverability of the scientific audiovisual media hosted in the portal. Text, speech, and image content of the video are automatically indexed by specialised GND (Gemeinsame Normdatei) subject headings. A semantic search is established based on properties of the GND ontology. The cross-lingual retrieval uses English 'translations' that were derived by an ontology mapping (DBpedia i. a.). Further ways of increasing the discoverability and reuse of the metadata are publishing them as Linked Open Data and interlinking them with other data sets.

Series

Communications in computer and information science; 544

0.003413712 = product of:
  0.047791965 = sum of:
    0.047791965 = weight(_text_:bibliothek in 1083) [ClassicSimilarity], result of:
      0.047791965 = score(doc=1083,freq=6.0), product of:
        0.121660605 = queryWeight, product of:
          4.1055303 = idf(docFreq=1980, maxDocs=44218)
          0.029633347 = queryNorm
        0.39283025 = fieldWeight in 1083, product of:
          2.4494898 = tf(freq=6.0), with freq of:
            6.0 = termFreq=6.0
          4.1055303 = idf(docFreq=1980, maxDocs=44218)
          0.0390625 = fieldNorm(doc=1083)
  0.071428575 = coord(1/14)

RSWK

Automatische Indexierung / Dokument / Wissenschaftliche Bibliothek

Subject

Automatische Indexierung / Dokument / Wissenschaftliche Bibliothek

0.0033447412 = product of:
  0.046826374 = sum of:
    0.046826374 = weight(_text_:bibliothek in 2470) [ClassicSimilarity], result of:
      0.046826374 = score(doc=2470,freq=4.0), product of:
        0.121660605 = queryWeight, product of:
          4.1055303 = idf(docFreq=1980, maxDocs=44218)
          0.029633347 = queryNorm
        0.38489348 = fieldWeight in 2470, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          4.1055303 = idf(docFreq=1980, maxDocs=44218)
          0.046875 = fieldNorm(doc=2470)
  0.071428575 = coord(1/14)

Content

Vgl.: https://journals.ub.uni-heidelberg.de/index.php/bibliothek/article/view/14022.

Source

Perspektive Bibliothek. 3(2014) H.1, S.85-110

0.0033447412 = product of:
  0.046826374 = sum of:
    0.046826374 = weight(_text_:bibliothek in 4595) [ClassicSimilarity], result of:
      0.046826374 = score(doc=4595,freq=4.0), product of:
        0.121660605 = queryWeight, product of:
          4.1055303 = idf(docFreq=1980, maxDocs=44218)
          0.029633347 = queryNorm
        0.38489348 = fieldWeight in 4595, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          4.1055303 = idf(docFreq=1980, maxDocs=44218)
          0.046875 = fieldNorm(doc=4595)
  0.071428575 = coord(1/14)

Content

Vgl.: http://journals.ub.uni-heidelberg.de/index.php/bibliothek/article/view/48394. Vgl. auch: URN (PDF): http://nbn-resolving.de/urn:nbn:de:bsz:16-pb-483949.

Source

Perspektive Bibliothek. 7(2018) H.1, S.33-54

0.0031590632 = product of:
  0.02211344 = sum of:
    0.0071344664 = weight(_text_:information in 905) [ClassicSimilarity], result of:
      0.0071344664 = score(doc=905,freq=4.0), product of:
        0.052020688 = queryWeight, product of:
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.029633347 = queryNorm
        0.13714671 = fieldWeight in 905, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.0390625 = fieldNorm(doc=905)
    0.014978974 = weight(_text_:retrieval in 905) [ClassicSimilarity], result of:
      0.014978974 = score(doc=905,freq=2.0), product of:
        0.08963835 = queryWeight, product of:
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.029633347 = queryNorm
        0.16710453 = fieldWeight in 905, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.0390625 = fieldNorm(doc=905)
  0.14285715 = coord(2/14)

Abstract

Der Artikel basiert auf einer Masterarbeit mit dem Titel "Automatische Indexierung in der sozialwissenschaftlichen Fachinformation. Eine Evaluationsstudie zur maschinellen Erschließung für die Datenbank SOLIS" (Kempf 2012), die im Rahmen des Aufbaustudiengangs Bibliotheks- und Informationswissenschaft an der Humboldt- Universität zu Berlin am Lehrstuhl Information Retrieval verfasst wurde. Auf der Grundlage des Schalenmodells zur Inhaltserschließung in der Fachinformation stellt der Artikel Evaluationsergebnisse eines automatischen Erschließungsverfahrens für den Einsatz in der sozialwissenschaftlichen Fachinformation vor. Ausgehend von dem von Krause beschriebenen Anwendungsszenario, wonach SOLIS-Datenbestände (Sozialwissenschaftliches Literaturinformationssystem) von geringerer Relevanz automatisch erschlossen werden sollten, wurden auf dieser Dokumentgrundlage zwei Testreihen mit der Indexierungssoftware MindServer der Firma Recommind durchgeführt. Neben den Auswirkungen allgemeiner Systemeinstellungen in der ersten Testreihe wurde in der zweiten Testreihe die Indexierungsleistung der Software für die Rand- und die Kernbereiche der Literaturdatenbank miteinander verglichen. Für letztere Testreihe wurden für beide Bereiche der Datenbank spezifische Versionen der Indexierungssoftware aufgebaut, die anhand von Dokumentkorpora aus den entsprechenden Bereichen trainiert wurden. Die Ergebnisse der Evaluation, die auf der Grundlage intellektuell generierter Vergleichsdaten erfolgt, weisen auf Unterschiede in der Indexierungsleistung zwischen Rand- und Kernbereichen hin, die einerseits gegen den Einsatz automatischer Indexierungsverfahren in den Randbereichen sprechen. Andererseits deutet sich an, dass sich die Indexierungsresultate durch den Aufbau fachteilgebietsspezifischer Trainingsmengen verbessern lassen.

Source

Information - Wissenschaft und Praxis. 64(2013) H.2/3, S.96-106

0.0031590632 = product of:
  0.02211344 = sum of:
    0.0071344664 = weight(_text_:information in 3151) [ClassicSimilarity], result of:
      0.0071344664 = score(doc=3151,freq=4.0), product of:
        0.052020688 = queryWeight, product of:
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.029633347 = queryNorm
        0.13714671 = fieldWeight in 3151, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.0390625 = fieldNorm(doc=3151)
    0.014978974 = weight(_text_:retrieval in 3151) [ClassicSimilarity], result of:
      0.014978974 = score(doc=3151,freq=2.0), product of:
        0.08963835 = queryWeight, product of:
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.029633347 = queryNorm
        0.16710453 = fieldWeight in 3151, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.0390625 = fieldNorm(doc=3151)
  0.14285715 = coord(2/14)

Abstract

Arabic news articles in electronic collections are difficult to study. Browsing by category is rarely supported. Although helpful machine-learning methods have been applied successfully to similar situations for English news articles, limited research has been completed to yield suitable solutions for Arabic news. In connection with a Qatar National Research Fund (QNRF)-funded project to build digital library community and infrastructure in Qatar, we developed software for browsing a collection of about 237,000 Arabic news articles, which should be applicable to other Arabic news collections. We designed a simple taxonomy for Arabic news stories that is suitable for the needs of Qatar and other nations, is compatible with the subject codes of the International Press Telecommunications Council, and was enhanced with the aid of a librarian expert as well as five Arabic-speaking volunteers. We developed tailored stemming (i.e., a new Arabic light stemmer called P-Stemmer) and automatic classification methods (the best being binary Support Vector Machines classifiers) to work with the taxonomy. Using evaluation techniques commonly used in the information retrieval community, including 10-fold cross-validation and the Wilcoxon signed-rank test, we showed that our approach to stemming and classification is superior to state-of-the-art techniques.

Source

Journal of the Association for Information Science and Technology. 67(2016) no.11, S.2667-2683

0.0031590632 = product of:
  0.02211344 = sum of:
    0.0071344664 = weight(_text_:information in 4005) [ClassicSimilarity], result of:
      0.0071344664 = score(doc=4005,freq=4.0), product of:
        0.052020688 = queryWeight, product of:
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.029633347 = queryNorm
        0.13714671 = fieldWeight in 4005, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.0390625 = fieldNorm(doc=4005)
    0.014978974 = weight(_text_:retrieval in 4005) [ClassicSimilarity], result of:
      0.014978974 = score(doc=4005,freq=2.0), product of:
        0.08963835 = queryWeight, product of:
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.029633347 = queryNorm
        0.16710453 = fieldWeight in 4005, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.0390625 = fieldNorm(doc=4005)
  0.14285715 = coord(2/14)

Abstract

Subject indexing is an intellectually intensive process that has many inherent uncertainties. Existing manual subject indexing systems generally produce binary outcomes for whether or not to assign an indexing term. This does not sufficiently reflect the extent to which the indexing terms are associated with the documents. On the other hand, the idea of probabilistic or weighted indexing was proposed a long time ago and has seen success in capturing uncertainties in the automatic indexing process. One hurdle to overcome in implementing weighted indexing in manual subject indexing systems is the practical burden that could be added to the already intensive indexing process. This study proposes a method to infer automatically the associations between subject terms and documents through text mining. By uncovering the connections between MeSH descriptors and document text, we are able to derive the weights of MeSH descriptors manually assigned to documents. Our initial results suggest that the inference method is feasible and promising. The study has practical implications for improving subject indexing practice and providing better support for information retrieval.

Source

Journal of the Association for Information Science and Technology. 66(2015) no.9, S.1776-1784

0.0028605436 = product of:
  0.020023804 = sum of:
    0.0050448296 = weight(_text_:information in 1518) [ClassicSimilarity], result of:
      0.0050448296 = score(doc=1518,freq=2.0), product of:
        0.052020688 = queryWeight, product of:
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.029633347 = queryNorm
        0.09697737 = fieldWeight in 1518, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.0390625 = fieldNorm(doc=1518)
    0.014978974 = weight(_text_:retrieval in 1518) [ClassicSimilarity], result of:
      0.014978974 = score(doc=1518,freq=2.0), product of:
        0.08963835 = queryWeight, product of:
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.029633347 = queryNorm
        0.16710453 = fieldWeight in 1518, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.0390625 = fieldNorm(doc=1518)
  0.14285715 = coord(2/14)

Content

Automatic Text Summarization Some Important Concepts 23 Single document Summarization 53 Guided Multi-Document Summarization 109 Emerging systems 151 Source and DomainSpecific Summarization 179 Text Abstracting 219 Evaluating Document Summaries 243 Conclusion 275 Information Retrieval NLP and Automatic Text Summarization 281 Automatic Text Summarization Resources 305

0.0028317936 = product of:
  0.019822555 = sum of:
    0.0049941265 = weight(_text_:information in 1054) [ClassicSimilarity], result of:
      0.0049941265 = score(doc=1054,freq=4.0), product of:
        0.052020688 = queryWeight, product of:
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.029633347 = queryNorm
        0.0960027 = fieldWeight in 1054, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.02734375 = fieldNorm(doc=1054)
    0.014828428 = weight(_text_:retrieval in 1054) [ClassicSimilarity], result of:
      0.014828428 = score(doc=1054,freq=4.0), product of:
        0.08963835 = queryWeight, product of:
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.029633347 = queryNorm
        0.16542503 = fieldWeight in 1054, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.02734375 = fieldNorm(doc=1054)
  0.14285715 = coord(2/14)

Abstract

Mit Hilfe eines algorithmisch arbeitenden Verfahrens können Mehrwortgruppen aus elektronisch vorliegenden Texten identifiziert und extrahiert werden. Als Datengrundlage für diese Arbeit dienen kunsthistorische Lexikonartikel des Reallexikons zur Deutschen Kunstgeschichte. Die linguistisch, wörterbuchbasierte Open-Source-Software Lingo wurde in dieser Studie genutzt. Mit Lingo ist es möglich, auf Basis erstellter Wortmuster, bestimmte Wortfolgen aus elektronisch vorliegenden Daten algorithmisch zu identifizieren und zu extrahieren. Die erstellten Wortmuster basieren auf Wortklassen, mit denen die lexikalisierten Einträge in den Wörterbüchern getaggt sind und dadurch näher definiert werden. So wurden individuelle Wortklassen für Fachterminologie, Eigennamen, oder Adjektive vergeben. In der vorliegenden Arbeit werden zusätzlich Funktionswörter in die Musterbildung mit einbezogen. Dafür wurden neue Wortklassen definiert. Funktionswörter bestimmen Artikel, Konjunktionen und Präpositionen. Ziel war es fachterminologische Mehrwortgruppen mit kunsthistorischen Inhalten zu extrahieren unter der gezielten Einbindung von Funktionswörtern. Anhand selbst gebildeter Kriterien, wurden die extrahierten Mehrwortgruppen qualitativ analysiert. Es konnte festgestellt werden, dass die Verwendung von Funktionswörtern fachterminologische Mehrwortgruppen erzeugt, die als potentielle Indexterme weitere Verwendung im Information Retrieval finden können.
Mehrwortgruppen sind als lexikalische Einheit zu betrachten und bestehen aus mindestens zwei miteinander in Verbindung stehenden Begriffen. Durch die Ver-bindung mehrerer Fachwörter transportieren sie in Fachtexten aussagekräftige Informationen. Sie vermitteln eindeutige Informationen, da aus den resultierenden Beziehungen zwischen den in Verbindung stehenden Fachbegriffen die inhaltliche Bedeutung eines Fachtextes ersichtlich wird. Demzufolge ist es sinnvoll, Mehrwort-gruppen aus Fachtexten zu extrahieren, da diese die Inhalte eindeutig repräsentieren. So können Mehrwortgruppen für eine inhaltliche Erschließung genutzt und beispiels-weise als Indexterme im Information Retrieval bereitgestellt werden. Mehrwortgruppen enthalten Informationen eines Textes, die in natürlicher Sprache vorliegen. Zur Extraktion von Informationen eines elektronisch vorliegenden Textes kommen maschinelle Verfahren zum Einsatz, da Sprache Strukturen aufweist, die maschinell verarbeitet werden können. Eine mögliche Methode Mehrwortgruppen innerhalb von elektronisch vorliegenden Fachtexten zu identifizieren und extrahieren ist ein algorithmisches Verfahren. Diese Methode erkennt Wortfolgen durch das Bilden von Wortmustern, aus denen sich eine Mehrwortgruppe in einem Text zusammensetzt. Die Wortmuster repräsentieren somit die einzelnen Bestandteile einer Mehrwortgruppe. Bereits an mathematischen Fachtexten wurde dieses Verfahren untersucht und analysiert. Relevante Mehrwortgruppen, die ein mathematisches Konzept oder mathe-matischen Inhalt repräsentierten, konnten erfolgreich extrahiert werden. Zum Einsatz kam das Indexierungssystem Lingo, mit dessen Programmodul sequencer eine algorithmische Identifizierung und Extraktion von Mehrwortgruppen möglich ist. In der vorliegenden Arbeit wird dieses algorithmische Verfahren unter Einsatz der Software Lingo genutzt, um Mehrwortgruppen aus kunsthistorischen Fachtexten zu extrahieren. Als Datenquelle dienen kunsthistorische Lexikonartikel aus dem Reallexikon zur Deutschen Kunstgeschichte, welches in deutscher Sprache vorliegt. Es wird untersucht, ob positive Ergebnisse im Sinne von fachterminologischen Mehrwort-gruppen mit kunsthistorischen Inhalten erzeugt werden können. Dabei soll zusätzlich die Einbindung von Funktionswörtern innerhalb einer Mehrwortgruppe erfolgen. Funktionswörter definieren Artikel, Konjunktionen und Präpositionen, die für sich alleinstehend keine inhaltstragende Bedeutung besitzen, allerdings innerhalb einer Mehrwortgruppe syntaktische Funktionen erfüllen. Anhand der daraus resultierenden Ergebnisse wird analysiert, ob das Hinzufügen von Funktionswörtern innerhalb einer Mehrwortgruppe zu positiven Ergebnissen führt. Ziel soll es demnach sein, fach-terminologische Mehrwortgruppen mit kunsthistorischen Inhalten zu erzeugen, unter Einbindung von Funktionswörtern. Bei der Extraktion fachterminologischer Mehrwortgruppen wird im Folgenden insbesondere auf die Erstellung von Wortmustern eingegangen, da diese die Basis liefern, mit welchen das Programmmodul sequencer Wortfolgen innerhalb der kunst-historischen Lexikonartikel identifiziert. Eine Einordung der Indexierungsergebnisse erfolgt anhand selbst gebildeter Kriterien, die definieren, was unter einer fach-terminologischen Mehrwortgruppe zu verstehen ist.

0.0019709074 = product of:
  0.027592704 = sum of:
    0.027592704 = weight(_text_:bibliothek in 2078) [ClassicSimilarity], result of:
      0.027592704 = score(doc=2078,freq=2.0), product of:
        0.121660605 = queryWeight, product of:
          4.1055303 = idf(docFreq=1980, maxDocs=44218)
          0.029633347 = queryNorm
        0.22680065 = fieldWeight in 2078, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          4.1055303 = idf(docFreq=1980, maxDocs=44218)
          0.0390625 = fieldNorm(doc=2078)
  0.071428575 = coord(1/14)

Content

Vgl. unter: http://edoc.hu-berlin.de/series/berliner-handreichungen/2010-284/PDF/284.pdf. Auch als: Automatische Indexierung von Dokumenten in einer wissenschaftlichen Bibliothek: Implementierung und Evaluierung am Beispiel der Deutschen Zentralbibliothek für Wirtschaftswissenschaften. Diplomica Verlag, 2011.

0.001815726 = product of:
  0.025420163 = sum of:
    0.025420163 = weight(_text_:retrieval in 1557) [ClassicSimilarity], result of:
      0.025420163 = score(doc=1557,freq=4.0), product of:
        0.08963835 = queryWeight, product of:
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.029633347 = queryNorm
        0.2835858 = fieldWeight in 1557, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.046875 = fieldNorm(doc=1557)
  0.071428575 = coord(1/14)

Abstract

In this paper, we present a multimodal Recurrent Neural Network (m-RNN) model for generating novel sentence descriptions to explain the content of images. It directly models the probability distribution of generating a word given previous words and the image. Image descriptions are generated by sampling from this distribution. The model consists of two sub-networks: a deep recurrent neural network for sentences and a deep convolutional network for images. These two sub-networks interact with each other in a multimodal layer to form the whole m-RNN model. The effectiveness of our model is validated on three benchmark datasets (IAPR TC-12 [8], Flickr 8K [28], and Flickr 30K [13]). Our model outperforms the state-of-the-art generative method. In addition, the m-RNN model can be applied to retrieval tasks for retrieving images or sentences, and achieves significant performance improvement over the state-of-the-art methods which directly optimize the ranking objective function for retrieval.

0.0017435154 = product of:
  0.024409214 = sum of:
    0.024409214 = weight(_text_:web in 1717) [ClassicSimilarity], result of:
      0.024409214 = score(doc=1717,freq=2.0), product of:
        0.09670874 = queryWeight, product of:
          3.2635105 = idf(docFreq=4597, maxDocs=44218)
          0.029633347 = queryNorm
        0.25239927 = fieldWeight in 1717, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.2635105 = idf(docFreq=4597, maxDocs=44218)
          0.0546875 = fieldNorm(doc=1717)
  0.071428575 = coord(1/14)

Content

Beitrag für die Tagung: Beyond libraries - subject metadata in the digital environment and semantic web. IFLA Satellite Post-Conference, 17-18 August 2012, Tallinn. Vgl.: http://http://www.nlib.ee/index.php?id=17763.

0.0017435154 = product of:
  0.024409214 = sum of:
    0.024409214 = weight(_text_:web in 1969) [ClassicSimilarity], result of:
      0.024409214 = score(doc=1969,freq=2.0), product of:
        0.09670874 = queryWeight, product of:
          3.2635105 = idf(docFreq=4597, maxDocs=44218)
          0.029633347 = queryNorm
        0.25239927 = fieldWeight in 1969, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.2635105 = idf(docFreq=4597, maxDocs=44218)
          0.0546875 = fieldNorm(doc=1969)
  0.071428575 = coord(1/14)

Footnote

Contribution in a special issue "Beyond libraries: Subject metadata in the digital environment and Semantic Web" - Enthält Beiträge der gleichnamigen IFLA Satellite Post-Conference, 17-18 August 2012, Tallinn.

0.0015801124 = product of:
  0.011060786 = sum of:
    0.005707573 = weight(_text_:information in 5499) [ClassicSimilarity], result of:
      0.005707573 = score(doc=5499,freq=4.0), product of:
        0.052020688 = queryWeight, product of:
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.029633347 = queryNorm
        0.10971737 = fieldWeight in 5499, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.03125 = fieldNorm(doc=5499)
    0.0053532133 = product of:
      0.016059639 = sum of:
        0.016059639 = weight(_text_:22 in 5499) [ClassicSimilarity], result of:
          0.016059639 = score(doc=5499,freq=2.0), product of:
            0.103770934 = queryWeight, product of:
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.029633347 = queryNorm
            0.15476047 = fieldWeight in 5499, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.03125 = fieldNorm(doc=5499)
      0.33333334 = coord(1/3)
  0.14285715 = coord(2/14)

Date

20. 1.2015 18:30:22

Footnote

Beitrag in einem Special Issue: Information Science in the German-speaking Countries.

Source

Aslib journal of information management. 71(2019) no.3, S.415-439

0.0014978976 = product of:
  0.020970564 = sum of:
    0.020970564 = weight(_text_:retrieval in 4310) [ClassicSimilarity], result of:
      0.020970564 = score(doc=4310,freq=2.0), product of:
        0.08963835 = queryWeight, product of:
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.029633347 = queryNorm
        0.23394634 = fieldWeight in 4310, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.0546875 = fieldNorm(doc=4310)
  0.071428575 = coord(1/14)

0.0012839122 = product of:
  0.01797477 = sum of:
    0.01797477 = weight(_text_:retrieval in 1868) [ClassicSimilarity], result of:
      0.01797477 = score(doc=1868,freq=2.0), product of:
        0.08963835 = queryWeight, product of:
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.029633347 = queryNorm
        0.20052543 = fieldWeight in 1868, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.046875 = fieldNorm(doc=1868)
  0.071428575 = coord(1/14)

Abstract

We present a model that generates free-form natural language descriptions of image regions. Our model leverages datasets of images and their sentence descriptions to learn about the inter-modal correspondences between text and visual data. Our approach is based on a novel combination of Convolutional Neural Networks over image regions, bidirectional Recurrent Neural Networks over sentences, and a structured objective that aligns the two modalities through a multimodal embedding. We then describe a Recurrent Neural Network architecture that uses the inferred alignments to learn to generate novel descriptions of image regions. We demonstrate the effectiveness of our alignment model with ranking experiments on Flickr8K, Flickr30K and COCO datasets, where we substantially improve on the state of the art. We then show that the sentences created by our generative model outperform retrieval baselines on the three aforementioned datasets and a new dataset of region-level annotations.

0.0012839122 = product of:
  0.01797477 = sum of:
    0.01797477 = weight(_text_:retrieval in 3622) [ClassicSimilarity], result of:
      0.01797477 = score(doc=3622,freq=2.0), product of:
        0.08963835 = queryWeight, product of:
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.029633347 = queryNorm
        0.20052543 = fieldWeight in 3622, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.024915 = idf(docFreq=5836, maxDocs=44218)
          0.046875 = fieldNorm(doc=3622)
  0.071428575 = coord(1/14)

Abstract

Indexing is contextualized and a brief description is provided of some of the most used automatic indexing systems. We describe SISA, a system which uses location heuristics rules, statistical rules like term frequency (TF) or TF-IDF to obtain automatic or semi-automatic indexing, depending on the user's preference. The aim of this research is to ascertain which rules (location heuristics rules or TF-IDF rules) provide the best indexing terms. SISA is used to obtain the automatic indexing of 200 scientific articles on fruit growing written in Portuguese. It uses, on the one hand, location heuristics rules founded on the value of certain parts of the articles for indexing such as titles, abstracts, keywords, headings, first paragraph, conclusions and references and, on the other, TF-IDF rules. The indexing is then evaluated to ascertain retrieval performance through recall, precision and f-measure. Automatic indexing of the articles with location heuristics rules provided the best results with the evaluation measures.

0.0011531039 = product of:
  0.016143454 = sum of:
    0.016143454 = weight(_text_:information in 442) [ClassicSimilarity], result of:
      0.016143454 = score(doc=442,freq=2.0), product of:
        0.052020688 = queryWeight, product of:
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.029633347 = queryNorm
        0.3103276 = fieldWeight in 442, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          1.7554779 = idf(docFreq=20772, maxDocs=44218)
          0.125 = fieldNorm(doc=442)
  0.071428575 = coord(1/14)

Source

Grundlagen der praktischen Information und Dokumentation. 6. Aufl