Global ETD Search

Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.

Results

Showing 1 to 20 of 28 for “"Rare words"”.

  1. Distributional initialization of neural networks

    … features of the input tokens (characters, words, phrases, etc.). Traditionally, for Neural Networks (NNs) such input representations are one-hot vectors, where each word is represented with a vector of all-but-one zeros, with value 1 on the position that corresponds to the index of the word …

    lmu-germany Repository record for Distributional initialization of neural networks (opens in a new tab)

  2. The Role of negative information when learning dense word vectors

    … other, mathematically formulated as two events (words and contexts) having negative Pointwise Mutual Information. We validate our hypothesis by creating an efficient factorization algorithm, LexVec, scalable to web-size corpora, that accounts for negative information in a principled way, closing …

    brazil-ufrgs Repository record for The Role of negative information when learning dense word vectors (opens in a new tab)

  3. Talking Down to Toddlers: Comparing Maternal Language to Adults, Maternal Language Input to Toddlers, and Toddler Vocabulary Growth

    … between ADS and CDS, 2) proportions of rare words, 3) one-word utterances and 4) nouns in CDS on toddler vocabularies are considered. Mothers' MLU and <italic>VOCD</italic> in CDS were stable, but did not relate to children's vocabulary growth, while other input factors were related to …

    maryland Repository record for Talking Down to Toddlers: Comparing Maternal Language to Adults, Maternal Language Input to Toddlers, and Toddler Vocabulary Growth (opens in a new tab)

  4. Character-level and syntax-level models for low-resource and multilingual natural language processing

    … and transliteration between Named Entities and rare words in different languages to improve the construction of Bilingual Word Embeddings (BWEs) and named entity resources, and (ii) exploit multiparallel corpora for projecting labels from high- to low-resource languages, thereby gaining access …

    lmu-germany Repository record for Character-level and syntax-level models for low-resource and multilingual natural language processing (opens in a new tab)

  5. Novel Methods for Natural Language Modeling and Pretraining

    … method leveraging the lexical feature for rare word modeling. It maps words that have a common WordNet hypernym to the same class and trains large neural LMs by gradually annealing from predicting the class to token prediction during training. Class-based prediction leads to an implicit …

    waterloo Repository record for Novel Methods for Natural Language Modeling and Pretraining (opens in a new tab)

  6. Multi-timescale representation learning in LSTM Language Models

    … Treebank and WikiText-2 datasets, especially on rare words.

    texas Repository record for Multi-timescale representation learning in LSTM Language Models (opens in a new tab)

  7. Realms of Strangers

    … in fifty episodic narratives to collect his rare words, sophisticated compositions, and curious accounts. A key element of the maqāma genre, as noted by Abū Isḥāq Ibrāhīm ibn ʿAlī al-Ḥuṣrī in the 5th/11th century, is its strangeness. Al-Ḥarīrī accentuates this element and employs it in two …

    fu-berlin Repository record for Realms of Strangers (opens in a new tab)

  8. Morph-based speech retrieval: Indexing methods and evaluations of unsupervised morphological analysis

    … (ASR) is used to transform the spoken words into text, and information retrieval (IR) methods are used for searching. Traditional ASR systems have a predefined vocabulary of words, and any word that is out-of-vocabulary (OOV) can not be recognized. Typically, rare words are excluded, …

    aalto Repository record for Morph-based speech retrieval: Indexing methods and evaluations of unsupervised morphological analysis (opens in a new tab)

  9. An Investigation into How Word Learning Modality Influences Subsequent Word Processing During Sentence Reading

    This dissertation executed a novel rare word learning paradigm to investigate whetherlearning words via different modalities (i.e., auditory vs. visual) influences subsequent word knowledge. The study was split across two days. On day one, adult participants learned the definitions or …

    alabama Repository record for An Investigation into How Word Learning Modality Influences Subsequent Word Processing During Sentence Reading (opens in a new tab)

  10. Memory for word presentations: The effects of word commonness and memorization strategy

    A list of words that occur frequently in everyday language is more recallable than a list of words that occur only rarely. This "word frequency" effect is eliminated, or even reversed slightly, if the common and rare words are mixed together in the same list. This pattern of results is replicated …

    rice Repository record for Memory for word presentations: The effects of word commonness and memorization strategy (opens in a new tab)

  11. Integrating Distributional, Compositional, and Relational Approaches to Neural Word Representations

    … input text sequences into space-delimited words, for which embeddings are trained over a large corpus by means of leveraging distributional information: a word is reducible to the set of contexts it appears in. This approach is powerful but imperfect; words not seen during the embedding …

    gatech Repository record for Integrating Distributional, Compositional, and Relational Approaches to Neural Word Representations (opens in a new tab)

  12. Adaptation, Context-aware Modeling and Rescoring Methods for Neural Language Models in Automatic Speech Recognition

    … and Transformer, to perform well on modeling rare words. Thus, in the first part of the dissertation, we focus on neural language model adaptation and contextual modeling methods. Specifically, we aim to develop general and lightweight approaches for real ASR systems. The second part of the …

    jhu Repository record for Adaptation, Context-aware Modeling and Rescoring Methods for Neural Language Models in Automatic Speech Recognition (opens in a new tab)

  13. Derivational morphology, spelling ability and dyslexia in secondary school students: assessment and intervention

    … structure on the spelling of derived words. A cross-sectional reading level-design was employed in order to determine differences between dyslexic adolescents and chronological-age and reading-age matched controls. The selection of the dyslexic participants was carried out through a …

    greece Repository record for Derivational morphology, spelling ability and dyslexia in secondary school students: assessment and intervention (opens in a new tab)

  14. Second chance competitive autoencoders for understanding textual data

    … track of this data, there are categories, keywords, tags, or labels that are assigned to each text. Dimensionality reduction and topic modeling in Mining text data has received a lot of attention. Topic modeling is a statistical technique for revealing the underlying semantic structure in a …

    umkc Repository record for Second chance competitive autoencoders for understanding textual data (opens in a new tab)

  15. Machine translation into morphologically rich low-resource languages

    … vocabularies, containing many poorly modeled rare words. This thesis addresses these challenges with multiple approaches. The focus is on methods for segmenting words into subwords, to get more frequent and thus easier learned representations, and to increase the symmetry between languages. It …

    aalto Repository record for Machine translation into morphologically rich low-resource languages (opens in a new tab)

  16. Setting letters and words into context

    … task. The optical imaging results showed that words elicit greater neural responses than nonwords in the left inferior parietal gyrus, which suggests a role of this region during the integration of orthographic, phonological and semantic representations. Greater activations for word stimuli in …

    fu-berlin Repository record for Setting letters and words into context (opens in a new tab)

  17. Η γλώσσα των δελφικών χρησμών: σχέσεις μαντικής και επικής γλώσσας στους χρησμούς του μαντείου από τον 8ο ως τον 5ο προχριστιανικό αιώνα

    … concerns the vocabulary (commonplace and rare words, formulae) and the metre and is based on formulaic analysis of verse first applied systematically in the Homeric epos by M. Parry. Oracular language has particular features such as: identical or slightly modified epic words in the same …

    greece Repository record for Η γλώσσα των δελφικών χρησμών: σχέσεις μαντικής και επικής γλώσσας στους χρησμούς του μαντείου από τον 8ο ως τον 5ο προχριστιανικό αιώνα (opens in a new tab)

  18. Statistical Learning for Structural Patterns with Trees

    … zero; and (2) many features that are counts of rarely occurring events are present, and appropriate aggregation of the rare features may lead to better estimation. In either case, the methods under consideration use a tree or a directed acyclic graph (DAG) that encodes relations among parameters …

    cornell Repository record for Statistical Learning for Structural Patterns with Trees (opens in a new tab)

  19. Advances in Spectral Learning with Applications to Text Analysis and Brain Imaging

    … -- low dimensional real vectors (called Eigenwords) that capture the “meaning” of words from their context. Second, we show how similar spectral methods can be applied to analyzing brain images. State-of-the-art approaches to learning word embeddings are slow to train or lack theoretical …

    penn Repository record for Advances in Spectral Learning with Applications to Text Analysis and Brain Imaging (opens in a new tab)

  20. Grade R teachers’ perceived competency and knowledge of oral language development in isiXhosa

    … however with the use of vocabulary and rare words, participants had challenges. Limitations in this case study included the diversity of languages which limited the scope to only nine schools that offered teaching and learning in isiXhosa throughout the district. This study cannot be …

    nwu-za Repository record for Grade R teachers’ perceived competency and knowledge of oral language development in isiXhosa (opens in a new tab)

Page 1 of 2