Search

Recommending products to users with intuitive explanations helps improve the system in transparency, persuasiveness, and satisfaction. Existing interpretation techniques include post hoc methods and interpretable modeling. The former category could quantitatively analyze input contribution to model prediction but has limited interpretation faithfulness, while the latter could explain model internal mechanisms but may not directly attribute model predictions to input features. In this study, we propose a novel Dual Interpretable Recommendation model called DIRECT, which integrates ideas of the two interpretation categories to inherit their advantages and avoid limitations. Specifically, DIRECT makes use of item descriptions as explainable evidence for recommendation. First, similar to the post hoc interpretation, DIRECT could attribute the prediction of a user preference score to textual words of the item descriptions. The attribution of each word is related to its sentiment polarity and word importance, where a word is important if it corresponds to an item aspect that the user is interested in. Second, to improve the interpretability of embedding space, we propose to extract high-level concepts from embeddings, where each concept corresponds to an item aspect. To learn discriminative concepts, we employ a concept bottleneck layer and maximize the coding rate reduction on word-aspect embeddings by leveraging a word–word affinity graph extracted from a pre-trained language model. In this way, DIRECT simultaneously achieves faithful attribution and usable interpretation of embedding space. We also show that DIRECT achieves linear inference time complexity regarding the length of item reviews. We conduct experiments including ablation studies on five real-world datasets. Quantitative analysis, visualizations, and case studies verify the interpretability of DIRECT. Our code is available at: https://github.com/JacksonWuxs/DIRECT. [ABSTRACT FROM AUTHOR]

Harnessing the power of Large Language Models (LLMs), this study explores the use of three state-of-the-art LLMs, specifically GPT-3.5-turbo, LLaMA3-8B, and LLaMA3-70B, for crash severity analysis and inference, framing it as a classification task. We generate textual narratives from original traffic crash tabular data using a pre-built template infused with domain knowledge. Additionally, we incorporated Chain-of-Thought (CoT) reasoning to guide the LLMs in analyzing the crash causes and then inferring the severity. This study also examine the impact of prompt engineering specifically designed for crash severity inference. The LLMs were tasked with crash severity inference to: (1) evaluate the models' capabilities in crash severity analysis, (2) assess the effectiveness of CoT and domain-informed prompt engineering, and (3) examine the reasoning abilities with the CoT framework. Our results showed that LLaMA3-70B consistently outperformed the other models, particularly in zero-shot settings. The CoT and Prompt Engineering techniques significantly enhanced performance, improving logical reasoning and addressing alignment issues. Notably, the CoT offers valuable insights into LLMs' reasoning process, unleashing their capacity to consider diverse factors such as environmental conditions, driver behavior, and vehicle characteristics in severity analysis and inference. [ABSTRACT FROM AUTHOR]

Large language models (LLMs) have demonstrated impressive capabilities in natural language processing. However, their internal mechanisms are still unclear and this lack of transparency poses unwanted risks for downstream applications. Therefore, understanding and explaining these models is crucial for elucidating their behaviors, limitations, and social impacts. In this article, we introduce a taxonomy of explainability techniques and provide a structured overview of methods for explaining Transformer-based language models. We categorize techniques based on the training paradigms of LLMs: traditional fine-tuning-based paradigm and prompting-based paradigm. For each paradigm, we summarize the goals and dominant approaches for generating local explanations of individual predictions and global explanations of overall model knowledge. We also discuss metrics for evaluating generated explanations and discuss how explanations can be leveraged to debug models and improve performance. Lastly, we examine key challenges and emerging opportunities for explanation techniques in the era of LLMs in comparison to conventional deep learning models. [ABSTRACT FROM AUTHOR]

Zero-shot relation triplet extraction (ZeroRTE) endeavors to extract relation triplets from a test set using a model trained on a training set with disjoint relations from the test set. Current ZeroRTE approaches primarily rely on two strategies: 1) Combining pre-trained language models to generate additional training samples; 2) Adding a large number of parameters that require training from scratch on top of a pre-trained language model. However, the former approach does not ensure the quality of generated samples, and the latter often struggles to generalize to unseen relations in the test set, particularly when the training set is small. In this paper, we introduce a novel method, N ext S entence P rediction for R elation T riplet E xtraction (NSP-RTE), abstracting ZeroRTE as a higher-level next sentence prediction (NSP) task to enhance its generalization ability to unseen relation categories. NSP-RTE integrates modules for relation recognition, entity detection, and triplet classification, leveraging pre-trained BERT models with fewer parameters requiring training from scratch, while eliminating the need for additional sample generation. Our experiments on the FewRel and Wiki-ZSL datasets demonstrate that NSP-RTE, with its simple and efficient design, significantly outperforms previous methods. • This paper transforms the zero-shot relation triplet extraction task into a next-sentence prediction problem for enhanced generalization. • The proposed method directly addresses the zero-shot relation triplet extraction task based on contextual semantics, eliminating the need for additional sample synthesis. • Experimental results on publicly available datasets demonstrate that the proposed method significantly outperforms previous methods. [ABSTRACT FROM AUTHOR]