In recent years, biomedical event extraction has been dominated by complicated pipeline and joint methods, which need to be simplified. In addition, existing work has not effectively utilized trigger word information ...
详细信息
ISBN:
(纸本)9789819794331;9789819794348
In recent years, biomedical event extraction has been dominated by complicated pipeline and joint methods, which need to be simplified. In addition, existing work has not effectively utilized trigger word information explicitly. Hence, we propose MLSL, a method based on multi-layer sequence labeling for joint biomedical event extraction. MLSL does not introduce prior knowledge and complex structures. Moreover, it explicitly incorporates the information of candidate trigger words into the sequence labeling to learn the interaction relationships between trigger words and argument roles. Based on this, MLSL can learn well with just a simple workflow. Extensive experimentation demonstrates the superiority of MLSL in terms of extraction performance compared to other state-of-the-art methods.
Even though large language models (LLMs) have demonstrated remarkable performance across various naturallanguageprocessing tasks, their application in speech-related tasks has largely remained underexplored. This wo...
详细信息
In this paper, we propose a fake news detection model that generates naturallanguage explanations and leverages these explanations for prediction. Deep learning models, particularly Large language Models like BERT, a...
详细信息
The rise of social networking services has increased text-based communication, often leading to misunderstandings. This study aims to develop a system using large language models (LLMs) like ChatGPT to provide real-ti...
详细信息
Social media anxiety detection refers to the approach of recognizing various signs of anxiety in a person based on their social media engagements. A person who experiences anxiety may express negative emotional respon...
详细信息
Social media has become the primary source of information for individuals, yet much of this information remains unverified. The rise of generative artificial intelligence has further accelerated the creation of unveri...
详细信息
ISBN:
(纸本)9789819794393;9789819794409
Social media has become the primary source of information for individuals, yet much of this information remains unverified. The rise of generative artificial intelligence has further accelerated the creation of unverified content. Adaptive rumor resolution systems are imperative for maintaining information integrity and public trust. Traditional methods have relied on encoder-based frameworks to enhance rumor representation and propagation characteristics. However, these models are often small in scale and lack generalizability for unforeseen events. Recent advances in Large language Models show promise but are unreliable in discerning truth from falsehood. Our work leverages LLMs by creating a testbed for predicting unprecedented rumors and designing a retrieval-augmented framework that integrates historical knowledge and collective intelligence. Experiments on two real-world datasets demonstrate the effectiveness of our proposed framework.
Navigating the dynamic hospitality landscape, premier hotels in India encounter a perpetual challenge: comprehending the intricate factors that drive guest satisfaction. While existing research provides broad insights...
详细信息
ISBN:
(纸本)9783031734762;9783031734779
Navigating the dynamic hospitality landscape, premier hotels in India encounter a perpetual challenge: comprehending the intricate factors that drive guest satisfaction. While existing research provides broad insights, the nuances of the luxury segment are often overlooked. This study aims to bridge this gap by meticulously analyzing reviews from over a hundred premier hotels in major metro cities in India, focusing on the 5-star category. The research identifies seven key attributes influencing reviewer perceptions and reveals distinct patterns. Food Quality, Hotel Staff, and Service emerge as powerful predictors of both positive and negative sentiments, while Cleanliness and Comfort exhibit unique impacts. Through the application of logistic regression and sentiment analysis, and naturallanguageprocessing (NLP) techniques such as tokenization, lemmatization, and VADER sentiment analysis on online reviews, this research quantifies the influence of each attribute. The findings contribute to a deeper understanding of guest expectations and provide actionable insights for hotel management to tailor offerings and enhance guest experiences. This study significantly contributes to the ongoing discourse on customer satisfaction in the hospitality sector in India.
In the medical field, unstructured medical text holds rich medical knowledge. Identifying medical entities in this text accurately is crucial for structured medical databases, knowledge graphs, and intelligent diagnos...
详细信息
ISBN:
(纸本)9789819794300;9789819794317
In the medical field, unstructured medical text holds rich medical knowledge. Identifying medical entities in this text accurately is crucial for structured medical databases, knowledge graphs, and intelligent diagnostic systems. Medical text has unique features, making it hard for traditional NER methods to identify complex medical entities. In particular, the recognition of nested entities within medical text poses a significant challenge, as it requires systems to recognize and understand the complex hierarchical relationships between entities, placing higher demands on traditional entity recognition systems. To overcome the challenges of nested entity recognition in medical text, we propose a method that combines semantic knowledge enhancement and global pointer optimization. Initially, we incorporate semantic prior knowledge of entity categories, capturing the interplay between labels and text by integrating label relationships. This allows us to obtain candidate entity information enriched with integrated label details. Following this, we establish a classification module to evaluate and score these candidate entities along with their labels, enabling entity prediction. To address nested entities, we introduce a Efficient GlobalPointer module that computes the likelihood of each text span being a specific entity type, thus bolstering nested entity recognition. By merging the outputs from both modules, we arrive at the final predicted entities. Experimental results indicate that our method excels on two flat entity datasets, CMedQANER and CCKS2017, as well as on the nested entity dataset CMeEE. Compared to baseline models, our approach demonstrates notable performance enhancements.
The rapid development of large-scale language models has garnered widespread interest from both academia and industry. Efficiently applying those models across various domains is now posing a challenge to researchers....
详细信息
ISBN:
(纸本)9789819794362;9789819794379
The rapid development of large-scale language models has garnered widespread interest from both academia and industry. Efficiently applying those models across various domains is now posing a challenge to researchers. High training costs and the relative scarcity of domain-specific data have rendered continual learning on general pretrained language models as one preferable approach. In this paper, we provide a comprehensive analysis and modification of these continual learning strategies for large language models, as they were initially designed for encoderonly architectures. Then a probing algorithm for the token representation shift was proposed to better alleviate forgetting. Additionally, corresponding evaluation metrics were modified for quantitative analysis of our methods. Through the experiment across three different domains, we verified the effectiveness of continual learning and probing algorithms on recent models. Results showed that knowledge distillation outperforms other methods in cross-domain continual learning. Moreover, the introduction of probing can further enhance the accuracy with a relatively small calculation budget.
Mathematical reasoning is challenging for large language models (LLMs), while the scaling relationship concerning LLM capacity is under-explored. Existing works have tried to leverage the rationales of LLMs to train s...
详细信息
ISBN:
(纸本)9789819794393;9789819794409
Mathematical reasoning is challenging for large language models (LLMs), while the scaling relationship concerning LLM capacity is under-explored. Existing works have tried to leverage the rationales of LLMs to train small language models (SLMs) for enhanced reasoning abilities, referred to as distillation. However, most existing distillation methods have not considered guiding the small models to solve problems progressively from simple to complex, which can be a more effective way. This study proposes a multi-step self questioning and answering (M-SQA) method that guides SLMs to solve complex problems by starting from simple ones. Initially, multi-step self-questioning and answering rationales are extracted from LLMs based on complexity-based prompting. Subsequently, these rationales are employed for distilling SLMs in a multi-task learning framework, during which the model learns to multi-step reason in a self questioning and answering way and answer each sub-question in a single step iteratively. Experiments on current mathematical reasoning tasks demonstrate the effectiveness of the proposed approach.
暂无评论