Not All Subjectivity Is the Same! Defining Desiderata for the Evaluation of Subjectivity in NLP

• URL: http://arxiv.org/abs/2603.28351v1
• Date: Mon, 30 Mar 2026 12:21:32 GMT
• Title: Not All Subjectivity Is the Same! Defining Desiderata for the Evaluation of Subjectivity in NLP
• Abstract summary: Subjective judgments are part of several NLP datasets.<n>Recent work is increasingly prioritizing models whose outputs reflect this diversity of perspectives.<n>It remains a question whether our evaluation practices align with these models' objectives.
• Score: 8.416822614574368
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Subjective judgments are part of several NLP datasets and recent work is increasingly prioritizing models whose outputs reflect this diversity of perspectives. Such responses allow us to shed light on minority voices, which are frequently marginalized or obscured by dominant perspectives. It remains a question whether our evaluation practices align with these models' objectives. This position paper proposes seven evaluation desiderata for subjectivity-sensitive models, rooted in how subjectivity is represented in NLP data and models. The desiderata are constructed in a top-down approach, keeping in mind the user-centric impact of such models. We scan the experimental setup of 60 papers and show that various aspects of subjectivity are still understudied: the distinction between ambiguous and polyphonic input, whether subjectivity is effectively expressed to the user, and a lack of interplay between different desiderata, amongst other gaps.

Related papers

• Perspectives in Play: A Multi-Perspective Approach for More Inclusive NLP Systems [3.011820285006942]
  This study proposes a new multi-perspective approach using soft labels to encourage the development of perspective aware models.<n>We conduct an analysis across diverse subjective text classification tasks, including hate speech, irony, abusive language, and stance detection.<n>Results show that the multi-perspective approach better approximates human label distributions, as measured by Jensen-Shannon Divergence (JSD)<n>Our approach exhibits lower confidence in tasks like irony and stance detection, likely due to the inherent subjectivity present in the texts.
  arXiv  Detail & Related papers  (2025-06-25T07:53:36Z)
• Embracing Diversity: A Multi-Perspective Approach with Soft Labels [3.529000007777341]
  We propose a new framework for designing perspective-aware models on stance detection task, in which multiple annotators assign stances based on a controversial topic.<n>Results show that the multi-perspective approach yields better classification performance (higher F1-scores)
  arXiv  Detail & Related papers  (2025-03-01T13:33:38Z)
• Diverging Preferences: When do Annotators Disagree and do Models Know? [92.24651142187989]
  We develop a taxonomy of disagreement sources spanning 10 categories across four high-level classes. We find that the majority of disagreements are in opposition with standard reward modeling approaches. We develop methods for identifying diverging preferences to mitigate their influence on evaluation and training.
  arXiv  Detail & Related papers  (2024-10-18T17:32:22Z)
• VHELM: A Holistic Evaluation of Vision Language Models [75.88987277686914]
  We present the Holistic Evaluation of Vision Language Models (VHELM) VHELM aggregates various datasets to cover one or more of the 9 aspects: visual perception, knowledge, reasoning, bias, fairness, multilinguality, robustness, toxicity, and safety. Our framework is designed to be lightweight and automatic so that evaluation runs are cheap and fast.
  arXiv  Detail & Related papers  (2024-10-09T17:46:34Z)
• VALOR-EVAL: Holistic Coverage and Faithfulness Evaluation of Large Vision-Language Models [57.43276586087863]
  Large Vision-Language Models (LVLMs) suffer from hallucination issues, wherein the models generate plausible-sounding but factually incorrect outputs. Existing benchmarks are often limited in scope, focusing mainly on object hallucinations. We introduce a multi-dimensional benchmark covering objects, attributes, and relations, with challenging images selected based on associative biases.
  arXiv  Detail & Related papers  (2024-04-22T04:49:22Z)
• Evaluating the Fairness of Discriminative Foundation Models in Computer Vision [51.176061115977774]
  We propose a novel taxonomy for bias evaluation of discriminative foundation models, such as Contrastive Language-Pretraining (CLIP) We then systematically evaluate existing methods for mitigating bias in these models with respect to our taxonomy. Specifically, we evaluate OpenAI's CLIP and OpenCLIP models for key applications, such as zero-shot classification, image retrieval and image captioning.
  arXiv  Detail & Related papers  (2023-10-18T10:32:39Z)
• SemEval-2023 Task 11: Learning With Disagreements (LeWiDi) [75.85548747729466]
  We report on the second edition of the LeWiDi series of shared tasks. This second edition attracted a wide array of participants resulting in 13 shared task submission papers.
  arXiv  Detail & Related papers  (2023-04-28T12:20:35Z)
• Towards Reliable Assessments of Demographic Disparities in Multi-Label Image Classifiers [11.973749734226852]
  We consider multi-label image classification and, specifically, object categorization tasks. Design choices and trade-offs for measurement involve more nuance than discussed in prior computer vision literature. We identify several design choices that look merely like implementation details but significantly impact the conclusions of assessments.
  arXiv  Detail & Related papers  (2023-02-16T20:34:54Z)
• In Search of Insights, Not Magic Bullets: Towards Demystification of the Model Selection Dilemma in Heterogeneous Treatment Effect Estimation [92.51773744318119]
  This paper empirically investigates the strengths and weaknesses of different model selection criteria. We highlight that there is a complex interplay between selection strategies, candidate estimators and the data used for comparing them.
  arXiv  Detail & Related papers  (2023-02-06T16:55:37Z)
• Considering user agreement in learning to predict the aesthetic quality [35.255447771350404]
  We propose a re-adapted multi-task attention network to predict both the mean opinion score and the standard deviation in an end-to-end manner. With such loss, the model is encouraged to learn the uncertainty of the content that is relevant to the diversity of observers' opinions. Experiments have demonstrated that the proposed multi-task aesthetic model achieves state-of-the-art performance on two different types of aesthetic datasets.
  arXiv  Detail & Related papers  (2021-10-13T18:00:36Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.