21 |
Modeling Sense Structure in Word Usage Graphs with the Weighted Stochastic Block Model ...
|
|
|
|
BASE
|
|
Show details
|
|
22 |
InFillmore: Frame-Guided Language Generation with Bidirectional Context ...
|
|
|
|
BASE
|
|
Show details
|
|
25 |
Warum wir so wenig über die Sprachen in Deutschland wissen: Spracheinstellungen als Erkenntnisbarriere
|
|
|
|
In: Diskurs Kindheits- und Jugendforschung / Discourse. Journal of Childhood and Adolescence Research ; 16 ; 4 ; 403-419 ; Perspektiven von Kindern und Jugendlichen auf sprachliche Diversität und Sprachbildungsprozesse (2021)
|
|
BASE
|
|
Show details
|
|
26 |
Essays on Representation Learning for Political Science Research
|
|
|
|
Abstract:
This dissertation consists of three papers about leveraging representation learning for political science research. Representation learning refers to techniques that learn a mapping between input data and a feature vector or tensor with respect to a task, such as classification or regression. These vectors or tensors capture abstract and relevant concepts in the data, making it easier to extract information. In the three papers, I show how representation learning allows political scientists to work with complex data such as text and images effectively. In the first paper, I propose using word embeddings to calculate partisan associations from Twitter users' bios. It only requires that some users in the corpus of tweets use partisan words in their bios. Intuitively, the word embeddings learn associations between non-partisan and partisan words from bios and extend those associations to all users. I apply the method to a collection of users who tweeted about election incidents during the 2016 United States general election. Which partisan accounts get retweeted, favorited, and followed, and which partisan hashtags are used closely correlate with the partisan association scores. I also apply the method to users who tweeted about masks during the COVID-19 pandemic. I find that users with more Democratic-leaning partisan association scores are more likely to use health advocacy hashtags, such as #MaskUp. In the second paper, I look at the automated classification of observations with both images and text. Most state-of-the-art vision-and-language models are unusable for most political science research, as they require all observations to have both image and text and require computationally expensive pretraining. This paper proposes a novel vision-and-language framework called multimodal representations using modality translation, or MARMOT. MARMOT presents two methodological contributions: it constructs representations for observations missing image or text, and it replaces computationally expensive pretraining with modality translation. Modality translation learns the patterns between images and their captions. MARMOT outperforms an ensemble text-only classifier in 19 of 20 categories in multilabel classifications of tweets reporting election incidents during the 2016 U.S. general election. MARMOT also shows significant improvements over the results of benchmark multimodal models on the Hateful Memes dataset, improving the best accuracy and area under the receiver operating characteristic curve (AUC) set by VisualBERT from 0.6473 to 0.6760 and 0.7141 to 0.7530, respectively. In the third paper, I turn to the issue of computationally studying language usage evolution over time. The corpora that political scientists typically work with are much smaller than the extensive corpora used in natural language processing research. Training a word embedding space over each period, the usual approach to studying language usage evolution, worsens the problem by splitting up the corpus into even smaller corpora. This paper proposes a framework that uses pretrained and non-pretrained embeddings to learn time-specific word embeddings, called the pretrained-augmented embeddings (PAE) framework. In the first application, I apply the PAE framework to a corpus of New York Times text data spanning several decades. The PAE framework matches human judgments of how specific words evolve in their usage much more closely than existing methods. In the second application, I apply the PAE framework to a corpus of tweets written during the COVID-19 pandemic about masking. The PAE framework automatically detects shifts in discussions about specific events during the COVID-19 pandemic vis-a-vis the keyword of interest. ; PHD ; Political Science ; University of Michigan, Horace H. Rackham School of Graduate Studies ; http://deepblue.lib.umich.edu/bitstream/2027.42/169642/1/pywu_1.pdf
|
|
Keyword:
computational social science; computer vision; multimodal machine learning; natural language processing; Political Science; representation learning; social media; Social Sciences; Statistics and Numeric Data
|
|
URL: https://doi.org/10.7302/2687 https://hdl.handle.net/2027.42/169642
|
|
BASE
|
|
Hide details
|
|
27 |
Vocabulário escrito de estudantes de escolas públicas do Rio Grande do Sul : um estudo léxico-estatístico
|
|
|
|
BASE
|
|
Show details
|
|
32 |
Caractériser un texte en français : les passages-clés des niveaux A1 et A2 du CECRL.
|
|
|
|
In: Actes des JADT 2020 ; JADT 2020 15èmes Journées internationales d’Analyse statistique des Données Textuelles ; https://hal.archives-ouvertes.fr/hal-02430322 ; JADT 2020 15èmes Journées internationales d’Analyse statistique des Données Textuelles, Jun 2020, Toulouse, France. 11 p ; https://jadt2020.sciencesconf.org/ (2020)
|
|
BASE
|
|
Show details
|
|
33 |
Identifying Language and Cognitive Profiles in Children With ASD via a Cluster Analysis Exploration: Implications for the New ICD-11
|
|
|
|
In: ISSN: 1939-3806 ; EISSN: 1939-3806 ; Autism Research ; https://hal.archives-ouvertes.fr/hal-02880841 ; Autism Research, International Society for Autism Research, Wiley Periodicals, Inc., 2020, ⟨10.1002/aur.2268⟩ (2020)
|
|
BASE
|
|
Show details
|
|
34 |
NTeALan Dictionaries Platforms: An Example Of Collaboration-Based Model
|
|
|
|
In: Proceedings of the 1st International Workshop on Language Technology Platforms (IWLTP 2020) ; https://hal.archives-ouvertes.fr/hal-02701912 ; Proceedings of the 1st International Workshop on Language Technology Platforms (IWLTP 2020), 2020, pp.11 - 16 (2020)
|
|
BASE
|
|
Show details
|
|
35 |
Standard-based Lexical Models for Automatically Structured Dictionaries ; Modèles lexicaux standardisés pour les dictionnaires à structure automatique
|
|
|
|
In: https://tel.archives-ouvertes.fr/tel-03153438 ; Computation and Language [cs.CL]. Université de Paris, 2020. English (2020)
|
|
BASE
|
|
Show details
|
|
36 |
Question Answering with Hybrid Data and Models ; Question-réponse utilisant des données et modèles hybrides
|
|
|
|
In: https://tel.archives-ouvertes.fr/tel-02890467 ; Document and Text Processing. Université Paris-Saclay, 2020. English. ⟨NNT : 2020UPASS024⟩ (2020)
|
|
BASE
|
|
Show details
|
|
37 |
It’s complicated! ; It’s complicated!: On Natural Language Processing Tools and Digital Humanities
|
|
|
|
In: “Tool Criticism 3.0” Workshop ; https://hal.archives-ouvertes.fr/hal-03084644 ; “Tool Criticism 3.0” Workshop, Jul 2020, Online (due to COVID, initially planned in Ottawa), Canada (2020)
|
|
BASE
|
|
Show details
|
|
38 |
Caractérisation de registres de langue par extraction de motifs séquentiels émergents
|
|
|
|
In: JADT 2020 : 15èmes Journées Internationales d'Analyse statistique des Données Textuelles ; https://hal.archives-ouvertes.fr/hal-03078450 ; JADT 2020 : 15èmes Journées Internationales d'Analyse statistique des Données Textuelles, Jun 2020, Toulouse, France (2020)
|
|
BASE
|
|
Show details
|
|
40 |
Categorizing Languages and Speakers: Processes of Erasure in Data Treatment and Presentation
|
|
Busch, Brigitta. - : Groupe de recherche diversité urbaine, 2020. : Érudit, 2020
|
|
BASE
|
|
Show details
|
|
|
|