Dominican Republic

Deep learning models have shown great success in question answering (QA), however, biases in the training data may lead to them amplifying or reflecting inequity. To probe for bias in QA systems, we create two benchmarks for closed and open domain question answering, consisting of ambiguous questions and bias metrics. We use these benchmarks with four QA models and find that open-domain QA models amplify biases more than their closed-domain counterparts, potentially due to a document multiplication effect of retriever models.

EMNLP 2021

Eliciting Bias in Question Answering Models through Ambiguity

EMNLP 2021 is planned to be a hybrid event in Punta Cana, Dominican Republic, with both on-site and fully virtual participation possible. The experience for on-site participants would closely approximate a normal pre-COVID *ACL conference, with 5-6 thematically organized parallel sessions and live Q/A and interactive discussion immediately after the talks. Presentations by virtual participants will be equitably interleaved with those of on-site participants, projected on the auditorium screens as if on-site, and also followed immediately by live Q/A and interactive discussion at a time during reasonable waking hours for the virtual presenter. For all participants, on-site and virtual, who are unable to attend a session due to either time-zone issues or because they are participating in another session live, talk recordings and slides will be available online at a minimum after the live presentation (and in many cases before as well), and questions may be submitted in advance on session-specific discussion boards and answered live in session with the usual visual aids if desired.

<iframe style="width:700px;height:400px" src="https://online.fliphtml5.com/ebtyf/ceby/"  seamless="seamless" scrolling="no" frameborder="0" allowtransparency="true" allowfullscreen="true" ></iframe>

Please Note: The EMNLP registration system is not currently connected to the underline site as we are still in the process of building out EMNLP 2021. You will receive access instructions from underline the week of November 1st. 

Access is given only to EMNLP upon registration, if you have not registered please do so [here](https://2021.emnlp.org/registration).

Registered attendees will receive access the week of November 1st.

EMNLP 2021 is planned to be a hybrid event in Punta Cana, Dominican Republic, with both on-site and fully virtual participation possible.

workshop paper

Recent state-of-the-art approaches in open-domain dialogue include training end-to-end deep-learning models to learn various conversational features like emotional content of response, symbolic transitions of dialogue contexts in a knowledge graph and persona of the agent and the user, among others. While neural models have shown reasonable results, modelling the cognitive processes that humans use when conversing with each other may improve the agent's quality of responses. A key element of natural conversation is to tailor one's response such that it accounts for concepts that the speaker and listener may or may not know and the contextual relevance of all prior concepts used in conversation. We show that a rich representation and explicit modeling of these psychological processes can improve predictions made by existing neural network models. In this work, we propose a novel probabilistic approach using Markov Random Fields (MRF) to augment existing deep-learning methods for improved next utterance prediction. Using human and automatic evaluations, we show that our augmentation approach significantly improves the performance of existing state-of-the-art retrieval models for open-domain conversational agents.

MRF-Chat: Improving Dialogue with Markov Random Fields

We introduce a new variation of the language modeling approach that uses schema-driven prompting to provide task-aware history encoding that is used for both categorical and non-categorical slots. We further improve performance by augmenting the prompting with schema descriptions, a naturally occurring source of in-domain knowledge. Our purely generative system achieves state-of-the-art performance onMultiWOZ 2.2 and achieves competitive performance on two other benchmarks: Multi-WOZ 2.1 and M2M.

Dialogue State Tracking with a Language Model using Schema-Driven Prompting

Humans often employ figurative language use in communication, including during interactions with dialog systems. Thus, it is important for real-world dialog systems to be able to handle popular figurative language constructs like metaphor and simile. In this work, we analyze the performance of existing dialog models in situations where the input dialog context exhibits use of figurative language. We observe large gaps in handling of figurative language when evaluating the models on two open domain dialog datasets. When faced with dialog contexts consisting of figurative language, some models show very large drops in performance compared to contexts without figurative language. We encourage future research in dialog modeling to separately analyze and report results on figurative language in order to better test model capabilities relevant to real-world use. Finally, we propose lightweight solutions to help existing models become more robust to figurative language by simply using an external resource to translate figurative language to literal (non-figurative) forms while preserving the meaning to the best extent possible.

Investigating Robustness of Dialog Models to Popular Figurative Language Constructs

Conversations aimed at determining good recommendations are iterative in nature. People often express their preferences in terms of a critique of the current recommendation (e.g.,“It doesn’t look good for a date”), requiring some degree of common sense for a preference to be inferred. In this work, we present a method for transforming a user critique into a positive preference (e.g., “I prefer more romantic”) in order to retrieve reviews pertaining to potentially better recommendations (e.g., “Perfect for a romantic dinner”). We leverage a large neural language model (LM) in a few-shot setting to perform critique-to-preference transformation, and we test two methods for retrieving recommendations: one that matches embeddings, and another that fine-tunes an LM for the task. We instantiate this approach in the restaurant domain and evaluate it using a new dataset of restaurant critiques. In an ablation study, we show that utilizing critique-to-preference transformation improves recommendations, and that there are at least three general cases that explain this improved performance.

It doesn't look good for a date: Transforming Critiques into Preferences for Conversational Recommendation Systems

Communicating with humans is challenging for AIs because it requires a shared understanding of the world, complex semantics (e.g., metaphors or analogies), and at times multi-modal gestures (e.g., pointing with a finger, or an arrow in a diagram). We investigate these challenges in the context of Iconary, a collaborative game of drawing and guessing based on Pictionary, that poses a novel challenge for the research community. In Iconary, a Guesser tries to identify a phrase that a Drawer is drawing by composing icons, and the Drawer iteratively revises the drawing to help the Guesser in response. This back-and-forth often uses canonical scenes, visual metaphor, or icon compositions to express challenging words, making it an ideal test for mixing language and visual/symbolic communication in AI. We propose models to play Iconary and train them on over 55,000 games between human players. Our models are skillful players and are able to employ world knowledge in language models to play with words unseen during training. Elite human players outperform our models, particularly at the drawing task, leaving an important gap for future research to address. We release our dataset, code, and evaluation setup as a challenge to the community at http://www.github.com/allenai/iconary.

Iconary: A Pictionary-Based Game for Testing Multimodal Communication with Drawings and Text

Multi-modal Retrieval of Tables and Texts Using Tri-encoder Models

In this paper, we study the possibility of unsupervised Multiple Choices Question Answering (MCQA). From very basic knowledge, the MCQA model knows that some choices have higher probabilities of being correct than others. The information, though very noisy, guides the training of an MCQA model. The proposed method is shown to outperform the baseline approaches on RACE and is even comparable with some supervised learning approaches on MC500.

Unsupervised Multiple Choices Question Answering: Start Learning from Basic Knowledge

A new full Book MRQA dataset, introducing the task of "Character Description Recognition".

GANDALF: a General Character Name Description Dataset for Long Fiction

Semantic Answer Similarity for Evaluating Question Answering Models

Dense retrieval has been shown to be effective for Open Domain Question Answering, surpassing sparse retrieval methods like BM25. One such model, REALM, (Guu et al., 2020) is an end-to-end dense retrieval system that uses MLM based pretraining for improved downstream QA performance. However, the current REALM setup uses limited resources and is not comparable in scale to more recent systems, contributing to its lower performance. Additionally, it relies on noisy supervision for retrieval during fine-tuning. We propose REALM++, where we improve upon the training and inference setups and introduce better supervision signal for improving performance, without any architectural changes. REALM++ achieves ~5.5% absolute accuracy gains over the baseline while being faster to train. It also matches the performance of large models which have 3x more parameters demonstrating the efficiency of our setup.

Eliciting Bias in Question Answering Models through Ambiguity

Next from EMNLP 2021

MRF-Chat: Improving Dialogue with Markov Random Fields

Similar lecture

Semantic Categorization of Social Knowledge for Commonsense Question Answering