More Like this

1. Using BERT Embeddings to Model Word Importance in Conversational Transcripts for Deaf and Hard of Hearing Users

   https://doi.org/10.18653/v1/2022.ltedi-1.5  

   Amin, Akhter Al; Hassan, Saad; Alm, Cecilia; Huenerfauth, Matt (January 2022, Proceedings of the Second Workshop on Language Technology for Equality, Diversity and Inclusion)

   Deaf and hard of hearing individuals regularly rely on captioning while watching live TV. Live TV captioning is evaluated by regulatory agencies using various caption evaluation metrics. However, caption evaluation metrics are often not informed by preferences of DHH users or how meaningful the captions are. There is a need to construct caption evaluation metrics that take the relative importance of words in transcript into account. We conducted correlation analysis between two types of word embeddings and human-annotated labelled word-importance scores in existing corpus. We found that normalized contextualized word embeddings generated using BERT correlated better with manually annotated importance scores than word2vec-based word embeddings. We make available a pairing of word embeddings and their human-annotated importance scores. We also provide proof-of-concept utility by training word importance models, achieving an F1-score of 0.57 in the 6-class word importance classification task. 

   more » « less
2. English‐learning infants developing sensitivity to vowel phonotactic cues to word segmentation

   https://doi.org/10.1111/desc.13564  

   Katsuda, Hironori; Sundara, Megha (September 2024, Developmental Science)

   Abstract Previous research has shown that when domain‐general transitional probability (TP) cues to word segmentation are in conflict with language‐specific stress cues, English‐learning 5‐ and 7‐month‐olds rely on TP, whereas 9‐month‐olds rely on stress. In two artificial languages, we evaluated English‐learning infants’ sensitivity to TP cues to word segmentation vis‐a‐vis language‐specific vowel phonotactic (VP) cues—English words do not end in lax vowels. These cues were either consistent or conflicting. When these cues were in conflict, 10‐month‐olds relied on the VP cues, whereas 5‐month‐olds relied on TP. These findings align with statistical bootstrapping accounts, where infants initially use domain‐general distributional information for word segmentation, and subsequently discover language‐specific patterns based on segmented words. Research HighlightsResearch indicates that when transitional probability (TP) conflicts with stress cues for word segmentation, English‐learning 9‐month‐olds rely on stress, whereas younger infants rely on TP.In two artificial languages, we evaluated English‐learning infants’ sensitivity to TP versus vowel phonotactic (VP) cues for word segmentation.When these cues conflicted, 10‐month‐olds relied on VPs, whereas 5‐month‐olds relied on TP.These findings align with statistical bootstrapping accounts, where infants first utilize domain‐general distributional information for word segmentation, and then identify language‐specific patterns from segmented words. 

   more » « less
3. Using LSTMs to Assess the Obligatoriness of Phonological Distinctive Features for Phonotactic Learning

   https://doi.org/10.18653/v1/P19-1155  

   Mirea, Nicole; Bicknell, Klinton (July 2019, Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics)

   To ascertain the importance of phonetic information in the form of phonological distinctive features for the purpose of segment-level phonotactic acquisition, we compare the performance of two recurrent neural network models of phonotactic learning: one that has access to distinctive features at the start of the learning process, and one that does not. Though the predictions of both models are significantly correlated with human judgments of non-words, the feature-naive model significantly outperforms the feature-aware one in terms of probability assigned to a held-out test set of English words, suggesting that distinctive features are not obligatory for learning phonotactic patterns at the segment level. 

   more » « less
4. Interactive Visualizations of Word Embeddings for K-12 Students

   Bandyopadhyay, Saptarashmi; Xu, Jason; Pawar, Neel; Touretzky, David (June 2022, Proceedings of the AAAI Conference on Artificial Intelligence)

   Word embeddings, which represent words as dense feature vectors, are widely used in natural language processing. In their seminal paper on word2vec, Mikolov and colleagues showed that a feature space created by training a word prediction network on a large text corpus will encode semantic information that supports analogy by vector arithmetic, e.g., "king" minus "man" plus "woman" equals "queen". To help novices appreciate this idea, people have sought effective graphical representations of word embeddings.We describe a new interactive tool for visually exploring word embeddings. Our tool allows users to define semantic dimensions by specifying opposed word pairs, e.g., gender is defined by pairs such as boy/girl and father/mother, and age by pairs such as father/son and mother/daughter. Words are plotted as points in a zoomable and rotatable 3D space, where the third ”residual” dimension encodes distance from the hyperplane defined by all the opposed word vectors with age and gender subtracted out. Our tool allows users to visualize vector analogies, drawing the vector from “king” to “man” and a parallel vector from “woman” to “king-man+woman”, which is closest to “queen”. Visually browsing the embedding space and experimenting with this tool can make word embeddings more intuitive. We include a series of experiments teachers can use to help K-12 students appreciate the strengths and limitations of this representation. 

   more » « less
5. Interactive Visualizations of Word Embeddings for K-12 Students

   https://doi.org/10.1609/aaai.v36i11.21548  

   Bandyopadhyay, Saptarashmi; Xu, Jason; Pawar, Neel; Touretzky, David (June 2022, Proceedings of the AAAI Conference on Artificial Intelligence)

   Word embeddings, which represent words as dense feature vectors, are widely used in natural language processing. In their seminal paper on word2vec, Mikolov and colleagues showed that a feature space created by training a word prediction network on a large text corpus will encode semantic information that supports analogy by vector arithmetic, e.g., "king" minus "man" plus "woman" equals "queen". To help novices appreciate this idea, people have sought effective graphical representations of word embeddings.We describe a new interactive tool for visually exploring word embeddings. Our tool allows users to define semantic dimensions by specifying opposed word pairs, e.g., gender is defined by pairs such as boy/girl and father/mother, and age by pairs such as father/son and mother/daughter. Words are plotted as points in a zoomable and rotatable 3D space, where the third ”residual” dimension encodes distance from the hyperplane defined by all the opposed word vectors with age and gender subtracted out. Our tool allows users to visualize vector analogies, drawing the vector from “king” to “man” and a parallel vector from “woman” to “king-man+woman”, which is closest to “queen”. Visually browsing the embedding space and experimenting with this tool can make word embeddings more intuitive. We include a series of experiments teachers can use to help K-12 students appreciate the strengths and limitations of this representation. 

   more » « less