A Fine-grained Interpretability Evaluation Benchmark for Neural NLP

Wang, Lijie; Shen, Yaozong; Peng, Shuyuan; Zhang, Shuai; Xiao, Xinyan; Líu, Hao; Tang, Hongxuan; Chen, Ying; Wang, Hua; Wang, Haifeng

doi:10.18653/v1/2022.conll-1.6

Cited by 8 publications

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A major bottleneck of interpretability studies is the availability of annotated benchmark datasets. In recent times, many interpretability evaluation benchmark datasets have been introduced for neural NLP tasks [15,81]. In this article, our objective is to extend this interpretability aspect toward the document ranking task.…”

Section: Interpretability Of Ranking Modelsmentioning

confidence: 99%

Extractive Explanations for Interpretable Text Ranking

Leonhardt

Rudra

Anand

2023

ACM Trans. Inf. Syst.

View full text Add to dashboard Cite

Neural document ranking models perform impressively well due to superior language understanding gained from pre-training tasks. However, due to their complexity and large number of parameters, these (typically transformer-based) models are often non-interpretable in that ranking decisions can not be clearly attributed to specific parts of the input documents. In this paper we propose ranking models that are inherently interpretable by generating explanations as a by-product of the prediction decision. We introduce the Select-And-Rank paradigm for document ranking, where we first output an explanation as a selected subset of sentences in a document. Thereafter, we solely use the explanation or selection to make the prediction, making explanations first-class citizens in the ranking process. Technically, we treat sentence selection as a latent variable trained jointly with the ranker from the final output. To that end, we propose an end-to-end training technique for Select-And-Rank models utilizing reparameterizable subset sampling using the Gumbel-max trick . We conduct extensive experiments to demonstrate that our approach is competitive to state-of-the-art methods. Our approach is broadly applicable to numerous ranking tasks and furthers the goal of building models that are interpretable by design . Finally, we present real-world applications that benefit from our sentence selection method.

show abstract

Section: Interpretability Of Ranking Modelsmentioning

confidence: 99%

Extractive Explanations for Interpretable Text Ranking

Leonhardt

Rudra

Anand

2023

ACM Trans. Inf. Syst.

View full text Add to dashboard Cite

show abstract

“…Without these segments, the original text would lose its offensive nature entirely. This line of research draws inspiration from previous Natural Language Processing (NLP) tasks focused on aspect-based sentiment analysis, which seeks to determine the sentiment conveyed in a text and identify the specific portions of text that express that sentiment through the use of attention-based deep neural network models [25,26].…”

Section: Offensive and Toxic Spans' Datasetsmentioning

confidence: 99%

Offensive Text Span Detection in Romanian Comments Using Large Language Models

Paraschiv,

Ion,

Dascalu

2023

Information

View full text Add to dashboard Cite

The advent of online platforms and services has revolutionized communication, enabling users to share opinions and ideas seamlessly. However, this convenience has also brought about a surge in offensive and harmful language across various communication mediums. In response, social platforms have turned to automated methods to identify offensive content. A critical research question emerges when investigating the role of specific text spans within comments in conveying offensive characteristics. This paper conducted a comprehensive investigation into detecting offensive text spans in Romanian language comments using Transformer encoders and Large Language Models (LLMs). We introduced an extensive dataset of 4800 Romanian comments annotated with offensive text spans. Moreover, we explored the impact of varying model sizes, architectures, and training data volumes on the performance of offensive text span detection, providing valuable insights for determining the optimal configuration. The results argue for the effectiveness of BERT pre-trained models for this span-detection task, showcasing their superior performance. We further investigated the impact of different sample-retrieval strategies for few-shot learning using LLMs based on vector text representations. The analysis highlights important insights and trade-offs in leveraging LLMs for offensive-language-detection tasks.

show abstract