Search Results for author: Malihe Alikhani

Found 46 papers, 28 papers with code

Signed Coreference Resolution

1 code implementation EMNLP 2021 Kayo Yin, Kenneth DeHaan, Malihe Alikhani

Coreference resolution is key to many natural language processing tasks and yet has been relatively unexplored in Sign Language Processing.

coreference-resolution

The Role of Context and Uncertainty in Shallow Discourse Parsing

no code implementations COLING 2022 Katherine Atwell, Remi Choi, Junyi Jessy Li, Malihe Alikhani

We find that including annotator accuracy and confidence improves model accuracy, and incorporating confidence in the model’s temperature function can lead to models with significantly better-calibrated confidence measures.

Discourse Parsing

Learning to Generate Context-Sensitive Backchannel Smiles for Embodied AI Agents with Applications in Mental Health Dialogues

1 code implementation13 Feb 2024 Maneesh Bilalpur, Mert Inan, Dorsa Zeinali, Jeffrey F. Cohn, Malihe Alikhani

To improve the rapport-building capabilities of embodied agents we annotated backchannel smiles in videos of intimate face-to-face conversations over topics such as mental health, illness, and relationships.

DisCGen: A Framework for Discourse-Informed Counterspeech Generation

1 code implementation29 Nov 2023 Sabit Hassan, Malihe Alikhani

In this work, we propose a novel framework based on theories of discourse to study the inferential links that connect counter speeches to the hateful comment.

Learning to Generate Equitable Text in Dialogue from Biased Training Data

1 code implementation10 Jul 2023 Anthony Sicilia, Malihe Alikhani

Absence of equitable and inclusive principles can hinder the formation of common ground, which in turn negatively impacts the overall performance of the system.

Decision Making Fairness +1

D-CALM: A Dynamic Clustering-based Active Learning Approach for Mitigating Bias

no code implementations26 May 2023 Sabit Hassan, Malihe Alikhani

While active learning (AL) has shown promise in training models with a small amount of annotated data, AL's reliance on the model's behavior for selective sampling can lead to an accumulation of unwanted bias rather than bias mitigation.

Active Learning Clustering +2

HumBEL: A Human-in-the-Loop Approach for Evaluating Demographic Factors of Language Models in Human-Machine Conversations

1 code implementation23 May 2023 Anthony Sicilia, Jennifer C. Gates, Malihe Alikhani

While demographic factors like age and gender change the way people talk, and in particular, the way people talk to machines, there is little investigation into how large pre-trained language models (LMs) can adapt to these changes.

Memorization

LEATHER: A Framework for Learning to Generate Human-like Text in Dialogue

3 code implementations14 Oct 2022 Anthony Sicilia, Malihe Alikhani

From this insight, we propose a new algorithm, and empirically, we demonstrate our proposal improves both task-success and human-likeness of the generated text.

Diversity Model Selection +1

APPDIA: A Discourse-aware Transformer-based Style Transfer Model for Offensive Social Media Conversations

1 code implementation COLING 2022 Katherine Atwell, Sabit Hassan, Malihe Alikhani

Then, we introduce the first discourse-aware style-transfer models that can effectively reduce offensiveness in Reddit text while preserving the meaning of the original text.

Style Transfer

How people talk about each other: Modeling Generalized Intergroup Bias and Emotion

2 code implementations14 Sep 2022 Venkata S Govindarajan, Katherine Atwell, Barea Sinno, Malihe Alikhani, David I. Beaver, Junyi Jessy Li

Current studies of bias in NLP rely mainly on identifying (unwanted or negative) bias towards a specific demographic group.

PANCETTA: Phoneme Aware Neural Completion to Elicit Tongue Twisters Automatically

no code implementations13 Sep 2022 Sedrick Scott Keh, Steven Y. Feng, Varun Gangal, Malihe Alikhani, Eduard Hovy

Through automatic and human evaluation, as well as qualitative analysis, we show that PANCETTA generates novel, phonetically difficult, fluent, and semantically meaningful tongue twisters.

Modeling Non-Cooperative Dialogue: Theoretical and Empirical Insights

1 code implementation15 Jul 2022 Anthony Sicilia, Tristan Maidment, Pat Healy, Malihe Alikhani

We use the tools of learning theory to develop a theoretical model for identifying non-cooperative interlocutors and apply this theory to analyze different communication strategies.

Learning Theory

PAC-Bayesian Domain Adaptation Bounds for Multiclass Learners

1 code implementation12 Jul 2022 Anthony Sicilia, Katherine Atwell, Malihe Alikhani, Seong Jae Hwang

Multiclass neural networks are a common tool in modern unsupervised domain adaptation, yet an appropriate theoretical description for their non-uniform sample complexity is lacking in the adaptation literature.

Unsupervised Domain Adaptation

Zero-shot Cross-Linguistic Learning of Event Semantics

no code implementations5 Jul 2022 Malihe Alikhani, Thomas Kober, Bashar Alhafni, Yue Chen, Mert Inan, Elizabeth Nielsen, Shahab Raji, Mark Steedman, Matthew Stone

Typologically diverse languages offer systems of lexical and grammatical aspect that allow speakers to focus on facets of event structure in ways that comport with the specific communicative setting and discourse constraints they face.

The Change that Matters in Discourse Parsing: Estimating the Impact of Domain Shift on Parser Error

4 code implementations Findings (ACL) 2022 Katherine Atwell, Anthony Sicilia, Seong Jae Hwang, Malihe Alikhani

Our results not only motivate our proposal and help us to understand its limitations, but also provide insight on the properties of discourse models and datasets which improve performance in domain adaptation.

Discourse Parsing Domain Adaptation +1

Modeling Intensification for Sign Language Generation: A Computational Approach

1 code implementation Findings (ACL) 2022 Mert İnan, Yang Zhong, Sabit Hassan, Lorna Quandt, Malihe Alikhani

To employ our strategies, we first annotate a subset of the benchmark PHOENIX-14T, a German Sign Language dataset, with different levels of intensification.

Text Generation

Including Facial Expressions in Contextual Embeddings for Sign Language Generation

no code implementations11 Feb 2022 Carla Viegas, Mert İnan, Lorna Quandt, Malihe Alikhani

State-of-the-art sign language generation frameworks lack expressivity and naturalness which is the result of only focusing manual signs, neglecting the affective, grammatical and semantic functions of facial expressions.

Text Generation

Words of Wisdom: Representational Harms in Learning From AI Communication

no code implementations16 Nov 2021 Amanda Buddemeyer, Erin Walker, Malihe Alikhani

Many educational technologies use artificial intelligence (AI) that presents generated or produced language to the learner.

Diversity Question Generation +1

Cross-Modal Coherence for Text-to-Image Retrieval

1 code implementation22 Sep 2021 Malihe Alikhani, Fangda Han, Hareesh Ravi, Mubbasir Kapadia, Vladimir Pavlovic, Matthew Stone

Common image-text joint understanding techniques presume that images and the associated text can universally be characterized by a single implicit model.

Image Retrieval Retrieval

Examining Covert Gender Bias: A Case Study in Turkish and English Machine Translation Models

1 code implementation INLG (ACL) 2021 Chloe Ciora, Nur Iren, Malihe Alikhani

As Machine Translation (MT) has become increasingly more powerful, accessible, and widespread, the potential for the perpetuation of bias has grown alongside its advances.

Machine Translation Translation

Including Signed Languages in Natural Language Processing

no code implementations ACL 2021 Kayo Yin, Amit Moryossef, Julie Hochgesang, Yoav Goldberg, Malihe Alikhani

Signed languages are the primary means of communication for many deaf and hard of hearing individuals.

NAREOR: The Narrative Reordering Problem

1 code implementation14 Apr 2021 Varun Gangal, Steven Y. Feng, Malihe Alikhani, Teruko Mitamura, Eduard Hovy

In this paper, we propose and investigate the task of Narrative Reordering (NAREOR) which involves rewriting a given story in a different narrative order while preserving its plot.

Aspectuality Across Genre: A Distributional Semantics Approach

no code implementations COLING 2020 Thomas Kober, Malihe Alikhani, Matthew Stone, Mark Steedman

The interpretation of the lexical aspect of verbs in English plays a crucial role for recognizing textual entailment and learning discourse-level inferences.

Natural Language Inference

Discourse Coherence, Reference Grounding and Goal Oriented Dialogue

no code implementations8 Jul 2020 Baber Khalid, Malihe Alikhani, Michael Fellner, Brian McMahan, Matthew Stone

Prior approaches to realizing mixed-initiative human--computer referential communication have adopted information-state or collaborative problem-solving approaches.

reinforcement-learning Reinforcement Learning (RL)

Achieving Common Ground in Multi-modal Dialogue

no code implementations ACL 2020 Malihe Alikhani, Matthew Stone

All communication aims at achieving common ground (grounding): interlocutors can work together effectively only with mutual beliefs about what the state of the world is, about what their goals are, and about how they plan to make their goals a reality.

Cross-modal Coherence Modeling for Caption Generation

no code implementations ACL 2020 Malihe Alikhani, Piyush Sharma, Shengjie Li, Radu Soricut, Matthew Stone

We use coherence relations inspired by computational models of discourse to study the information needs and goals of image captioning.

Caption Generation controllable image captioning +1

Clue: Cross-modal Coherence Modeling for Caption Generation

no code implementations2 May 2020 Malihe Alikhani, Piyush Sharma, Shengjie Li, Radu Soricut, Matthew Stone

We use coherence relations inspired by computational models of discourse to study the information needs and goals of image captioning.

Caption Generation controllable image captioning +1

That and There: Judging the Intent of Pointing Actions with Robotic Arms

1 code implementation13 Dec 2019 Malihe Alikhani, Baber Khalid, Rahul Shome, Chaitanya Mitash, Kostas Bekris, Matthew Stone

This work proposes a set of interpretive principles for how a robotic arm can use pointing actions to communicate task information to people by extending existing models from the related literature.

Common Sense Reasoning

AI2D-RST: A multimodal corpus of 1000 primary school science diagrams

no code implementations9 Dec 2019 Tuomo Hiippala, Malihe Alikhani, Jonas Haverinen, Timo Kalliokoski, Evanfiya Logacheva, Serafina Orekhova, Aino Tuomainen, Matthew Stone, John A. Bateman

This article introduces AI2D-RST, a multimodal corpus of 1000 English-language diagrams that represent topics in primary school natural sciences, such as food webs, life cycles, moon phases and human physiology.

Question Answering Visual Question Answering

``Caption'' as a Coherence Relation: Evidence and Implications

no code implementations WS 2019 Malihe Alikhani, Matthew Stone

We study verbs in image{--}text corpora, contrasting \textit{caption} corpora, where texts are explicitly written to characterize image content, with \textit{depiction} corpora, where texts and images may stand in more general relations.

Diversity Image Retrieval +2

CITE: A Corpus of Image-Text Discourse Relations

1 code implementation NAACL 2019 Malihe Alikhani, Sreyasi Nag Chowdhury, Gerard de Melo, Matthew Stone

This paper presents a novel crowd-sourced resource for multimodal discourse: our resource characterizes inferences in image-text contexts in the domain of cooking recipes in the form of coherence relations.

Common Sense Reasoning

Cannot find the paper you are looking for? You can Submit a new open access paper.