• Publications
  • Influence
Modeling Empathy and Distress in Reaction to News Stories
TLDR
This contribution presents the first publicly available gold standard for empathy prediction, constructed using a novel annotation methodology which reliably captures empathy assessments by the writer of a statement using multi-item scales. Expand
Unsupervised Post-processing of Word Vectors via Conceptor Negation
TLDR
A novel word vector post-processing technique based on matrix conceptors (Jaeger2014), a family of regularized identity maps that can be used for the downstream natural language processing task of dialogue state tracking, yielding improved results in different dialogue domains. Expand
Comparison of Diverse Decoding Methods from Conditional Language Models
TLDR
This work performs an extensive survey of decoding-time strategies for generating diverse outputs from a conditional language model, and presents a novel method where over-sample candidates, then use clustering to remove similar sequences, thus achieving high diversity without sacrificing quality. Expand
Continual Learning for Sentence Representations Using Conceptors
TLDR
This paper proposes to initialize sentence encoders with the help of corpus-independent features, and then sequentially update sentence encodes using Boolean operations of conceptor matrices to learn corpus-dependent features. Expand
Conceptor Debiasing of Word Representations Evaluated on WEAT
TLDR
It is shown that conceptor debiasing diminishes racial and gender bias of word representations as measured using the Word Embedding Association Test (WEAT) of Caliskan et al. (2017). Expand
Complexity-Weighted Loss and Diverse Reranking for Sentence Simplification
TLDR
This work incorporates content word complexities, as predicted with a leveled word complexity model, into the loss function during training and generates a large set of diverse candidate simplifications at test time, and rerank these to promote fluency, adequacy, and simplicity. Expand
ChatEval: A Tool for Chatbot Evaluation
TLDR
A unified framework for human evaluation of chatbots that augments existing tools and provides a web-based hub for researchers to share and compare their dialog systems and open-source baseline models and evaluation datasets are introduced. Expand
Domain Aware Neural Dialog System
TLDR
DOM-Seq2Seq is presented, a domain aware neural network model based on the novel technique of using domain-targeted sequence-to-sequence models and a domain classifier to facilitate the formation of relevant responses in a conversation comprising of different domains. Expand
Predicting Emotional Word Ratings using Distributional Representations and Signed Clustering
TLDR
This work develops a method that automatically extends word-level ratings to unrated words using signed clustering of vector space word representations along with affect ratings, which achieves superior out-of-sample word rating prediction on both affective dimensions across three different languages when compared to state- of-the-art word similarity based methods. Expand
Collecting Verified COVID-19 Question Answer Pairs
TLDR
A dataset of over 2,100 COVID19 related Frequently asked Question-Answer pairs scraped from over 40 trusted websites is released and an additional 24, 000 questions pulled from online sources that have been aligned by experts with existing answered questions from this dataset are included. Expand
...
1
2
3
4
5
...