| 1 |
Matching Tweets With Applicable Fact-Checks Across Languages ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 3 |
Deep Learning for Text Style Transfer: A Survey
|
|
|
|
In: Computational Linguistics, 48 (1) (2022)
|
|
BASE
|
|
Show details
|
|
|
| 4 |
FIBER: Fill-in-the-Blanks as a Challenging Video Understanding Evaluation Framework ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 5 |
Exploring Self-Identified Counseling Expertise in Online Support Forums ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 6 |
Room to Grow: Understanding Personal Characteristics Behind Self Improvement Using Social Media ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 7 |
Analyzing the Surprising Variability in Word Embedding Stability Across Languages ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 8 |
STaCK: Sentence Ordering with Temporal Commonsense Knowledge ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 9 |
Hitting your MARQ: Multimodal ARgument Quality Assessment in Long Debate Video ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 10 |
How Good Is NLP? A Sober Look at NLP Tasks through the Lens of Social Impact ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 15 |
How Good Is NLP?A Sober Look at NLP Tasks through the Lens of Social Impact ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 16 |
How Good Is NLP?A Sober Look at NLP Tasks through the Lens of Social Impact
|
|
|
|
In: Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021 (2021)
|
|
BASE
|
|
Show details
|
|
|
| 17 |
Leveraging Longitudinal Data for Personalized Prediction and Word Representations
|
|
|
|
Abstract:
This thesis focuses on personalization, word representations, and longitudinal dialog. We first look at users expressions of individual preferences. In this targeted sentiment task, we find that we can improve entity extraction and sentiment classification using domain lexicons and linear term weighting. This task is important to personalization and dialog systems, as targets need to be identified in conversation and personal preferences affect how the system should react. Then we examine individuals with large amounts of personal conversational data in order to better predict what people will say. We consider extra-linguistic features that can be used to predict behavior and to predict the relationship between interlocutors. We show that these features improve over just using message content and that training on personal data leads to much better performance than training on a sample from all other users. We look not just at using personal data for these end-tasks, but also constructing personalized word representations. When we have a lot of data for an individual, we create personalized word embeddings that improve performance on language modeling and authorship attribution. When we have limited data, but we have user demographics, we can instead construct demographic word embeddings. We show that these representations improve language modeling and word association performance. When we do not have demographic information, we show that using a small amount of data from an individual, we can calculate similarity to existing users and interpolate or leverage data from these users to improve language modeling performance. Using these types of personalized word representations, we are able to provide insight into what words vary more across users and demographics. The kind of personalized representations that we introduce in this work allow for applications such as predictive typing, style transfer, and dialog systems. Importantly, they also have the potential to enable more equitable language models, with improved performance for those demographic groups that have little representation in the data. ; PHD ; Computer Science & Engineering ; University of Michigan, Horace H. Rackham School of Graduate Studies ; http://deepblue.lib.umich.edu/bitstream/2027.42/167971/1/cfwelch_1.pdf
|
|
Keyword:
Computer Science; Engineering; longitudinal dialog; natural language processing; personalization; Science
|
|
URL: https://hdl.handle.net/2027.42/167971
https://doi.org/10.7302/1398
|
|
BASE
|
|
Hide details
|
|
|
| 18 |
Leveraging Social Media as a Thermometer to Gauge Patient and Caregiver Concerns: COVID-19 and Prostate Cancer
|
|
|
|
In: Eur Urol Open Sci (2021)
|
|
BASE
|
|
Show details
|
|
|
| 19 |
"Judge me by my size (noun), do you?'' YodaLib: A Demographic-Aware Humor Generation Framework ...
|
|
|
|
BASE
|
|
Show details
|
|
|
|
|
