United Arab Emirates

Task-Oriented Dialogue (TOD) systems are drawing more and more attention in recent studies. Current methods focus on constructing pre-trained models or fine-tuning strategies while the evaluation of TOD is limited by a policy mismatch problem. That is, during evaluation, the user utterances are from the annotated dataset while these utterances should interact with previous responses which can have many alternatives besides annotated texts. Therefore, in this work, we propose an interactive evaluation framework for TOD. We first build a goaloriented user simulator based on pre-trained models and then use the user simulator to interact with the dialogue system to generate dialogues. Besides, we introduce a sentencelevel and a session-level score to measure the sentence fluency and session coherence in the interactive evaluation. Experimental results show that RL-based TOD systems trained by our proposed user simulator can achieve nearly 98% inform and success rates in the interactive evaluation of MultiWOZ dataset and the proposed scores measure the response quality besides the inform and success rates. We are hoping that our work will encourage simulatorbased interactive evaluations in the TOD task.

EMNLP 2022

Is MultiWOZ a Solved Task? An Interactive TOD Evaluation Framework with User Simulator

task-oriented-dialogue

usersimulator

evaluation

## Welcome to EMNLP 2022!
I am delighted to welcome you to EMNLP 2022! I believe this conference has been complicated beyond any precedent. Over the past year, it’s been thrilling to see the organization team approach each new puzzle with creativity and enthusiasm. We hope that those participating in Abu Dhabi as well as those joining remotely will leave the conference feeling newly inspired by the program and newly connected to our ever-growing community. Following EMNLP 2021 and major NLP conferences since, EMNLP 2022 is “hybrid,” serving both virtual and in-person participants.

Our key innovations for EMNLP 2022 include:

* EMNLP 2022 is “hybrid” in a second sense, as well: we allowed both direct and rolling review paper submissions, building on the pilot experiment of EMNLP 2021, which considered a small number of ARR submissions. 
* Familiar from NAACL but new to EMNLP, we’ve added an industry track.
* During the conference, “portals” will link virtual poster sessions to in-person conference participants during poster sessions each day.
* The first ACL-family conference in the United Arab Emirates.

 *Message from Noah A. Smith, University of Washington and Allen Institute for AI, Seattle, Washington, USA* 
***EMNLP 2022 General Chair***
 
[![](https://assets.underline.io/uploads/markdown_image/1/image/9eec7d4a287ee18c278b08229290aa83.png)](https://drive.google.com/file/d/1OlPv6QBeo62VVTughj2jkiLeyHd1WnUt/view)
 
[![](https://assets.underline.io/uploads/markdown_image/1/image/a3db7a768409f05192210d98601edb25.png)](https://emnlp2022.rocket.chat/)

To access this site you need to register. Please register [here](https://2022.emnlp.org/registration/).

Register here

Welcome!
EMNLP 2022 will take place in Abu Dhabi from December 7th to December 11th, 2022. And it will be held in hybrid mode, both online and offline.

workshop paper

In the real world, autonomous driving agents navigate in highly dynamic environments full of unexpected situations where pre-trained models are unreliable. In these situations, what is immediately available to vehicles is often only human operators. Empowering autonomous driving agents with the ability to navigate in a continuous and dynamic environment and to communicate with humans through sensorimotor-grounded dialogue becomes critical. To this end, we introduce Dialogue On the ROad To Handle Irregular Events (DOROTHIE), a novel interactive simulation platform that enables the creation of unexpected situations on the fly to support empirical studies on situated communication with autonomous driving agents. Based on this platform, we created the Situated Dialogue Navigation (SDN), a navigation benchmark of 183 trials with a total of 8415 utterances, around 18.7 hours of control streams and 2.9 hours of trimmed audio. SDN is developed to evaluate the agent’s ability to predict dialogue moves from humans as well as generate its own dialogue moves and physical navigation actions. We further developed a transformer-based baseline model for these SDN tasks. Our empirical results indicate that language guided-navigation in a highly dynamic environment is an extremely difficult task for end-to-end models. These results will provide insight towards future work on robust autonomous driving agents.

DOROTHIE: Spoken Dialogue for Handling Unexpected Situations in Interactive Autonomous Driving Agents

It is still a pipe dream that personal AI assistants on the phone and AR glasses can assist our daily life in addressing our questions like “how to adjust the date for this watch?” and “how to set its heating duration? (while pointing at an oven)”. The queries used in conventional tasks (i.e. Video Question Answering, Video Retrieval, Moment Localization) are often factoid and based on pure text. In contrast, we present a new task called Taskoriented Question-driven Video Segment Retrieval (TQVSR). Each of our questions is an image-box-text query that focuses on affordance of items in our daily life and expects relevant answer segments to be retrieved from a corpus of instructional video-transcript segments. To support the study of this TQVSR task, we construct a new dataset called AssistSR. We design novel guidelines to create high-quality samples. This dataset contains 3.2k multimodal questions on 1.6k video segments from instructional videos on diverse daily-used items. To address TQVSR, we develop a simple yet effective model called Dual Multimodal Encoders (DME) that significantly outperforms several baseline methods while still having large room for improvement in the future. Moreover, we present detailed ablation analyses. Code and data are available at https//github.com/StanLei52/TQVSR.

AssistSR: Task-oriented Video Segment Retrieval for Personal AI Assistant

Visual grounding of Language aims at enriching textual representations of language with multiple sources of visual knowledge such as images and videos. Although visual grounding is an area of intense research, inter-lingual aspects of visual grounding have not received much attention. The present study investigates the inter-lingual visual grounding of word embeddings. We propose an implicit alignment technique between the two spaces of vision and language in which inter-lingual textual information interacts in order to enrich pre-trained textual word embeddings. We focus on three languages in our experiments, namely, English, Arabic, and German. We obtained visually grounded vector representations for these languages and studied whether visual grounding on one or multiple languages improved the performance of embeddings on word similarity and categorization benchmarks. Our experiments suggest that inter-lingual knowledge improves the performance of grounded embeddings in similar languages such as German and English. However, inter-lingual grounding of German or English with Arabic led to a slight degradation in performance on word similarity benchmarks. On the other hand, we observed an opposite trend on categorization benchmarks where Arabic had the most improvement on English. In the discussion section, several reasons for those findings are laid out. We hope that our experiments provide a baseline for further research on inter lingual visual grounding.

Visual Grounding of Inter-lingual Word-Embeddings

Discourse relations are typically modeled as a discrete class that characterizes the relation between segments of text (e.g. causal explanations, expansions). However, such predefined discrete classes limit the universe of potential relationships and their nuanced differences. Adding higher-level semantic structure to contextual word embeddings, we propose representing discourse relations as points in high dimensional continuous space. However, unlike words, discourse relations often have no surface form (relations are in between two segments, often with no word or phrase in that gap) which presents a challenge for existing embedding techniques. We present a novel method for automatically creating discourse relation embeddings (DiscRE), addressing the embedding challenge through a weakly supervised, multitask approach to learn diverse and nuanced relations in social media. Results show DiscRE representations obtain the best performance on Twitter discourse relation classification (macro F1=0.76), social media causality prediction (from F1=0.79 to 0.81), and perform beyond modern sentence and word transformers at traditional discourse relation classification, capturing novel nuanced relations (e.g. relations at the intersection of causal explanations and counterfactuals).

Discourse Relation Embeddings: Representing the Relations between Discourse Segments in Social Media

Image captioning models tend to describe images in an object-centric way, emphasising visible objects. But image descriptions can also abstract away from objects and describe the type of scene depicted. In this paper, we explore the potential of a state of the art Vision and Language model, VinVL, to caption images at the scene level using  (1) a novel dataset which pairs images with both object-centric and scene descriptions. Through (2) an in-depth analysis of the effect of the fine-tuning, we show (3) that a small amount of curated data suffices to generate scene descriptions without losing the capability to identify object-level concepts in the scene; the model acquires a more holistic view of the image compared to when object-centric descriptions are generated. We discuss the parallels between these results and insights from computational and cognitive science research on scene perception.

Understanding Cross-modal Interactions in V&L Models that Generate Scene Descriptions

We explore the utility of explicit structure for representation learning in NLP by developing StrAE： an autoencoding framework that faithfully leverages given sentence structure to learn multi-level node embeddings in an unsupervised fashion, which is then evaluated on a suite of downstream tasks. We train StrAE using four different input structures and multiple objectives, including a novel contrastive objective over structure. We find that using explicit structure can lead to better embeddings for downstream tasks compared to a non-structural baseline (GloVE) trained on the same data. Also, the kind of structure used has a measurable effect on task performance, with structure induced from the attention of a large language model (RoBERTa) performing best. Moreover, the right kind of structure employed with the contrastive objective performs on par with a Fasttext baseline on averaged downstream task performance, without access to subword information. Finally, averaged across different tasks and kinds of structure, the novel contrastive objective outperforms standard cross-entropy.

StrAE: Autoencoding for Pre-Trained Embeddings using Explicit Structure

This paper introduces DeepParliament, a legal domain Benchmark Dataset that gathers bill documents and metadata and performs various bill status classification tasks. The proposed dataset text covers a broad range of bills from 1986 to the present and contains richer information on parliament bill content. Data collection, detailed statistics and analyses are provided in the paper. Moreover, we experimented with different types of models ranging from RNN to pretrained and reported the results. We are proposing two new benchmarks: Binary and Multi-Class Bill Status classification. Models developed for bill documents and relevant supportive tasks may assist Members of Parliament (MPs), presidents, and other legal practitioners. It will help review or prioritise bills, thus speeding up the billing process, improving the quality of decisions and reducing the time consumption in both houses. Considering that the foundation of the country"s democracy is Parliament and state legislatures, we anticipate that our research will be an essential addition to the Legal NLP community. This work will be the first to present a Parliament bill prediction task. In order to  improve the accessibility of legal AI resources and promote reproducibility, we have made our code and dataset publicly accessible at github.com/monk1337/DeepParliament.

DeepParliament: A Legal domain Benchmark & Dataset for Parliament Bills Prediction

Automatic Speech Recognition (ASR) systems are taking over in different industries starting from producing video subtitles to interactive digital assistants. ASR output can be used in automatic indexing, categorizing, searching along with normal human readability. Raw transcripts from ASR systems are difficult to interpret since it usually produces text without punctuation and case information (all lower, all upper, camel case etc.), thus limiting the performance of downstream NLP tasks. We proposed an approach to restore the punctuation and case for both English and Hinglish (i.e Hindi vocabulary in Latin script) languages. We have performed a classification task using encoder-based transformers which is a mini BERT consisting of 4 encoder layers for punctuation and case restoration instead of the traditional Seq2Seq model considering the latency constraint in real world use cases. It consists of  a total number of 15 distinct classes for the model which includes 5 punctuations i.e Period(.), Comma(,), Single Quote(‘), Double Quote(") & Question Mark(?) with different combinations of casing. The model is benchmarked on an internal dataset which was based on user conversation with the voice assistant and it achieves a F1(macro) score of 91.52% on the test set.

Punctuation and case restoration in code mixed Indian languages

Adversarial attack of structured prediction models faces various challenges such as the difficulty of perturbing discrete words, the sentence quality issue, and the sensitivity of outputs to small perturbations.  In this work, we introduce SHARP, a new attack method that formulates the black-box adversarial attack as a search-based optimization problem with a specially designed objective function considering sentence fluency, meaning preservation and attacking effectiveness. Additionally, three different searching strategies are analyzed and compared, \ie, Beam Search, Metropolis-Hastings Sampling, and Hybrid Search. We demonstrate the effectiveness of our attacking strategies on two challenging structured prediction tasks: part-of-speech (POS) tagging and dependency parsing. Through automatic and human evaluations, we show that our method performs a more potent attack compared with pioneer arts. Moreover, the generated adversarial examples can be used to successfully boost the robustness and performance of the victim model via adversarial training.

Probing Script Knowledge from Pre-Trained Models

This paper describes the Inria ALMAnaCH team submission to the WMT 2022 general translation shared task. Participating in the language directions \{cs,ru,uk\}→en and cs↔uk, we experiment with the use of a dedicated Latin-script transcription convention aimed at representing all Slavic languages involved in a way that maximises character- and word-level correspondences between them as well as with the English language. Our hypothesis was that bringing the source and target language closer could have a positive impact on machine translation results. We provide multiple comparisons, including bilingual and multilingual baselines, with and without transcription. Initial results indicate that the transcription strategy was not successful, resulting in lower results than baselines. We nevertheless submitted our multilingual, transcribed models as our primary systems, and in this paper provide some indications as to why we got these negative results.

Downloads

Next from EMNLP 2022

DOROTHIE: Spoken Dialogue for Handling Unexpected Situations in Interactive Autonomous Driving Agents

Similar lecture

Is MultiWOZ a Solved Task? An Interactive TOD Evaluation Framework with User Simulator