Dominican Republic

Learning authors representations from their textual productions is now widely used to solve multiple downstream tasks, such as classification, link prediction or user recommendation. Author embedding methods are often built on top of either Doc2Vec (Mikolov et al. 2014) or the Transformer architecture (Devlin et al. 2019). Evaluating the quality of these embeddings and what they capture is a difficult task. Most articles use either classification accuracy or authorship attribution, which does not clearly measure the quality of the representation space, if it really captures what it has been built for. In this paper, we propose a novel evaluation framework of author embedding methods based on the writing style. It allows to quantify if the embedding space effectively captures a set of stylistic features, chosen to be the best proxy of an author writing style. This approach gives less importance to the topics conveyed by the documents. It turns out that recent models are mostly driven by the inner semantic of authors&#39; production. They are outperformed by simple baselines, based on state-of-the-art pretrained sentence embedding models, on several linguistic axes. These baselines can grasp complex linguistic phenomena and writing style more efficiently, paving the way for designing new style-driven author embedding models.

EMNLP 2021

Writing Style Author Embedding Evaluation

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.

Learning authors representations from their textual productions is now widely used to solve multiple downstream tasks, such as classification, link prediction or user recommendation. Author embedding methods are often built on top of either Doc2Vec (Mikolov et al. 2014) or the Transformer architecture (Devlin et al. 2019). Evaluating the quality of these embeddings and what they capture is a difficult task. Most articles use either classification accuracy or authorship attribution, which does not clearly measure the quality of the representation space, if it really captures what it has been built for. In this paper, we propose a novel evaluation framework of author embedding methods based on the writing style. It allows to quantify if the embedding space effectively captures a set of stylistic features, chosen to be the best proxy of an author writing style. This approach gives less importance to the topics conveyed by the documents. It turns out that recent models are mostly driven by the inner semantic of authors' production. They are outperformed by simple baselines, based on state-of-the-art pretrained sentence embedding models, on several linguistic axes. These baselines can grasp complex linguistic phenomena and writing style more efficiently, paving the way for designing new style-driven author embedding models.

workshop paper

Code-mixed language plays a crucial role in communication in multilingual societies. Though the recent growth of web users has greatly boosted the use of such mixed languages, the current generation of dialog systems is primarily monolingual. This increase in the usage of code-mixed language has prompted dialog systems in a similar language. We present our work in Code-Mixed Dialog Generation, an unexplored task in code-mixed languages, generating utterances in code-mixed language rather than a single language that is more often just English. We present a new synthetic corpus in code-mix for dialogs, CM-DailyDialog, by converting an existing English-only dialog corpus to a mixed Hindi-English corpus. We then propose a baseline approach where we show the effectiveness of using mBART like multilingual sequence-to-sequence transformers for code-mixed dialog generation. Our best performing dialog models can conduct coherent conversations in Hindi-English mixed language as evaluated by human and automatic metrics setting new benchmarks for the Code-Mixed Dialog Generation task.

Towards Code-Mixed Hinglish Dialogue Generation

Recently neural response generation models have leveraged large pre-trained transformer models and knowledge snippets to generate relevant and informative responses. However, this does not guarantee that generated responses are factually correct. In this paper, we examine factual correctness in knowledgegrounded neural response generation models. We present a human annotation setup to identify three different response types: responses that are factually consistent with respect to the input knowledge, responses that contain hallucinated knowledge, and non-verifiable chitchat style responses. We use this setup to annotate responses generated using different stateof-the-art models, knowledge snippets, and decoding strategies. In addition, to facilitate the development of a factual consistency detector, we automatically create a new corpus called Conv-FEVER that is adapted from the Wizard of Wikipedia dataset and includes factually consistent and inconsistent responses. We demonstrate the benefit of our Conv-FEVER dataset by showing that the models trained on this data perform reasonably well to detect factually inconsistent responses with respect to the provided knowledge through evaluation on our human annotated data. We will release the Conv-FEVER dataset and the human annotated responses.

Rome was built in 1776: A Case Study on Factual Correctness in Knowledge-Grounded Response Generation

Most prior work on task-oriented dialogue systems is restricted to supporting domain APIs. However, users may have requests that are out of the scope of these APIs. This work focuses on identifying such user requests. Existing methods for this task mainly rely on fine-tuning pre-trained models on large annotated data. We propose a novel method, REDE, based on adaptive representation learning and density estimation. REDE can be applied to zero-shot cases, and quickly learns a high-performing detector with only a few shots by updating less than 3K parameters. We demonstrate REDE's competitive performance on DSTC9 data and our newly collected test set.

Towards Zero and Few-shot Knowledge-seeking Turn Detection in Task-orientated Dialogue Systems

An interactive instruction following task has been proposed as a benchmark for learning to map natural language instructions and first-person vision into sequences of actions to interact with objects in a 3D simulated environment.
We find that an existing end-to-end neural model for this task is not robust to variations of objects and language instructions.
We assume that this problem is due to the high sensitiveness of neural feature extraction to small changes in vision and language inputs.
To mitigate this problem, we propose a neuro-symbolic approach that performs reasoning over high-level symbolic representations that are robust to small changes in raw inputs.
Our experiments on the ALFRED dataset show that our approach significantly outperforms the existing model by 18, 52, and 73 points in the success rate on the ToggleObject, PickupObject, and SliceObject subtasks in unseen environments respectively.

Improving the Robustness to Variations of Objects and Instructions with A Neuro-Symbolic Approach for Interactive Instruction Following

Swarm Robotics, which studies collective behaviors of large populations of interacting robots with simple embodied cognition, is an ideal testbed for studying the cultural evolution of language. Our goal is to expand existing simulations in order to explore a dominant theory in evolutionary linguistics—the self-domestication hypothesis—which suggests that increased prosociality led to the sophistication of the cultural niche, enabling the complexification of languages. In our model, robots in multiple nests are engaged in a foraging task (i.e., gathering resources in their environment) while playing a language game. Crucially, we include two novel features: (1) robot individuation: robots have a partner-specific memory, keeping track of the outcomes of past interactions with specific robots; (2) parameterizable prosociality: robots’ tendency to interact is based on experience: successful communication between robots reduces their aggression toward each other and increases their chances of interacting again. First, we show that these manipulations lead to the formation of a classic “in-group bias” where robots favor interaction with some robots over others - a bias which is highly common in social animals in nature but that was so far absent from swarm robotics models. Second, we show that prosociality provides an evolutionary advantage, with higher prosociality values resulting in the collection of more resources. Finally, we show that prosociality modulates the effect of physical distance on lexical convergence, such that low values of innate prosocially lead to more stable sub-swarm divergence, even in relative proximity (i.e., different nests robustly converge on different lexical variants despite being close to each other).

First Steps Toward a Swarm Robotics Model of Self-Domestication and Language Evolution

Machine-in-the-loop writing aims to enable humans to collaborate with models to effectively complete their writing tasks. Prior work in the creative domain has found that providing humans with a machine written draft or sentence level continuations has limited success as the generated text tends to deviate from the human's intentions. We train a rewriting model that, when prompted, modifies targeted spans of text within the user's original draft, enabling the human to retain control over the content while still taking advantage of the strengths of text generation models to introduce descriptive and figurative elements locally in the text. We evaluate the model on it's ability to collaborate with humans on the task of creative image captioning through a user study on Amazon Mechanical Turk. Users report that the model is helpful and third-party evaluation shows that users write more descriptive and figurative captions on average in the collaborative setting compared to a baseline of the human completing the task alone.

Machine-in-the-Loop Rewriting for Creative Image Captioning

Language-guided robots performing home and office tasks must navigate in and interact with the world. Grounding language instructions against visual observations and actions to take in an environment is an open challenge. We present Embodied BERT (EmBERT), a transformer-based model which can attend to high-dimensional, multi-modal inputs across long temporal horizons for language-conditioned task completion. Additionally, we bridge the gap between successful object-centric navigation models used for non-interactive agents and the language-guided visual task completion benchmark, ALFRED, by introducing object navigation targets for EmBERT training. EmBERT achieves competitive performance on the ALFRED benchmark, and is the first model to use a full, pretrained BERT stack while handling the long-horizon, dense, multi-modal histories of ALFRED.

Embodied BERT: A Transformer Model for Embodied, Language-guided Visual Task Completion

Detecting lexical semantic change in smaller data sets, e.g. in historical linguistics and digital humanities, is challenging due to a lack of statistical power. This issue is exacerbated by non-contextual embedding models that produce one embedding per word and, therefore, mask the variability present in the data. In this article, we propose an approach to estimate semantic shift by combining contextual word embeddings with permutation-based statistical tests. We use the false discovery rate procedure to address the large number of hypothesis tests being conducted simultaneously. We demonstrate the performance of this approach in simulation where it achieves consistently high precision by suppressing false positives. We additionally analyze real-world data from SemEval-2020 Task 1 and the Liverpool FC subreddit corpus. We show that by taking sample variation into account, we can improve the robustness of individual semantic shift estimates without degrading overall performance.

Statistically Significant Detection of Semantic Shifts using Contextual Word Embeddings

We present a collection of diverse datasets and aggregated scores of dataset aspects to help develop more generalizable findings in authorship attribution research.

Developing a Benchmark for Reducing Data Bias in Authorship Attribution

We propose a new reference-free summary quality evaluation measure, with emphasis on the faithfulness. The measure is based on finding and counting all probable potential inconsistencies of the summary with respect to the source document. The proposed ESTIME, Estimator of Summary-to-Text Inconsistency by Mismatched Embeddings, correlates with expert scores in summary-level SummEval dataset stronger than other common evaluation measures not only in Consistency but also in Fluency. We also introduce a method of generating subtle factual errors in human summaries. We show that ESTIME is more sensitive to subtle errors than other common evaluation measures.

Content not yet available

Writing Style Author Embedding Evaluation

Next from EMNLP 2021

Towards Code-Mixed Hinglish Dialogue Generation

Similar lecture

A Large-Scale Dataset for Empathetic Response Generation

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES