arXiv.itsnick.co

an independent field guide to 23 million papers from arXiv & OpenAlex — not affiliated with arXiv.org

Author

Lin Pan

185 papers · 2002–2026

2002: 32003: 12004: 12005: 72006: 62007: 22008: 12009: 82010: 102011: 42012: 112013: 82014: 72015: 62016: 32017: 62018: 62019: 102020: 172021: 152022: 62023: 112024: 172025: 92026: 10
2002peak 17 in 20202026

Frequent co-authors

Papers

showing 1–25 of 185 RSS

Improved Text Classification via Contrastive Adversarial Training

Lin Pan, Chung-Wei Hang, Avirup Sil +1 more

cs.CL · · cited 123× · arXiv ↗ · more like this ·

We propose a simple and general method to regularize the fine-tuning of Transformer-based encoders for text classification tasks. Specifically, during fine-tuning we generate adversarial examples by perturbing the word embeddings of the model and perform contrastive learning on clean and adversarial examples in order to teach the model to learn noise-invariant representations. By training on both clean and adversarial examples along with the additional contrastive objective, we observe consistent improvement over standard fine-tuning on clean examples. On several GLUE benchmark tasks, our fine-tuned BERT Large model outperforms BERT Large baseline by 1.7% on average, and our fine-tuned RoBERTa Large improves over RoBERTa Large baseline by 1.3%. We additionally validate our method in different domains using three intent classification datasets, where our fine-tuned RoBERTa Large…

Media Frames across Stages of Health Crisis: A Crisis Management Approach to News Coverage of Flu Pandemic

Po‐Lin Pan, Juan Meng

Communication · Journal of Contingencies and Crisis Management · · cited 96× · DOI ↗ · more like this ·

Dividing a crisis management process into three stages, this study examined news coverage of swine flu crisis in terms of news frames, mortality exemplars, vaccine problems, evaluation approaches to risk magnitudes and news sources. Results revealed that various framing strategies were used in news media at different stages of the flu crisis. The frames of health risk, societal problems, political/legal issues and prevention/health education were mostly used at the pre‐crisis stage, while the medical/scientific frame was regularly utilized at the post‐crisis stage to highlight medical treatment and scientific research. The evaluation approaches were also employed differently. The qualitative approach was mostly used at the pre‐crisis stage, while the quantitative and statistical approaches were applied frequently at the post‐crisis stage. Health professionals were widely cited as news…

How does destination crisis event type impact tourist emotion and forgiveness? The moderating role of destination crisis history

Lujun Su, Lin Pan, Yinghua Huang

Communication · Tourism Management · · cited 76× · DOI ↗ · more like this ·

Drawing on attribution theory and situational crisis communication theory, this study investigates how destination crisis events impact tourist sympathy, anger, and intentions of forgiveness in four experiments. It also examines the moderating effects of destination crisis history (none vs. similar vs. dissimilar). The results of Study 1 and Study 2 revealed that external crisis events cause more sympathy and tourist forgiveness than internal ones, but less anger. They also indicated that sympathy and anger play a mediating role in destination crisis events and tourist forgiveness. The results of Study 3 and Study 4 revealed that destination crisis history predicts the impact of crisis events on tourist emotion and forgiveness. In particular, when there is no destination crisis history or similar crisis history, an external crisis event will garner more sympathy and forgiveness than an…

Effects of tourism experiences on tourists’ subjective well-being through recollection and storytelling

Lujun Su, Lin Pan, Jun Wen +1 more

Sociology and Political Science · Journal Of Vacation Marketing · · cited 76× · DOI ↗ · more like this ·

In this study, we construct an integrated model demonstrating how the tourism experience (TE) affects tourists’ subjective well-being (SWB) via recollection and storytelling. We distributed questionnaires to tourists in two destinations in China (Kulangsu and Dr Sun Yat-Sen’s Mausoleum) and gathered 320 and 510 valid responses, respectively. The empirical results of Study 1 (Kulangsu) indicated that the tourism experience (TE) directly promoted tourists’ recollection, storytelling, and SWB. Additionally, recollection directly affected storytelling but not SWB. Storytelling further positively affected SWB. Mediation results revealed the following paths: 1) tourism experiences affected storytelling, which influenced SWB; 2) tourism experiences affected recollection, which influenced storytelling and in turn SWB. Findings from Study 2 (Dr Sun Yat-Sen’s Mausoleum) corroborated those of…

An Improved Apriori-based Algorithm for Association Rules Mining

Huan Wu, Zhigang Lü, Lin Pan +2 more

Information Systems · · cited 68× · DOI ↗ · more like this ·

Because of the rapid growth in worldwide information, efficiency of association rules mining (ARM) has been concerned for several years. In this paper, based on the original Apriori algorithm, an improved algorithm IAA is proposed. IAA adopts a new count-based method to prune candidate itemsets and uses generation record to reduce total data scan amount. Experiments demonstrate that our algorithm outperforms the original Apriori and some other existing ARM methods.

Morality or equality? Ideological framing in news coverage of gay marriage legitimization

Po‐Lin Pan, Juan Meng, Shuhua Zhou

Communication · The Social Science Journal · · cited 67× · DOI ↗ · more like this ·

This content analytic study investigated the approaches of two mainstream newspapers—The New York Times and the Chicago Tribune—to cover the gay marriage issue. The study used the Massachusetts legitimization of gay marriage as a dividing point to look at what kinds of specific political or social topics related to gay marriage were highlighted in the news media. The study examined how news sources were framed in the coverage of gay marriage, based upon the newspapers’ perspectives and ideologies. The results indicated that The New York Times was inclined to emphasize the topic of human equality related to the legitimization of gay marriage. After the legitimization, The New York Times became an activist for gay marriage. Alternatively, the Chicago Tribune highlighted the importance of human morality associated with the gay marriage debate. The perspective of the Chicago Tribune was not…

Importance of Synthesizing High-quality Data for Text-to-SQL Parsing

Yiyun Zhao, Jiarong Jiang, Yiqun Hu +14 more

cs.CL · · cited 54× · arXiv ↗ · more like this ·

Recently, there has been increasing interest in synthesizing data to improve downstream text-to-SQL tasks. In this paper, we first examined the existing synthesized datasets and discovered that state-of-the-art text-to-SQL algorithms did not further improve on popular benchmarks when trained with augmented synthetic data. We observed two shortcomings: illogical synthetic SQL queries from independent column sampling and arbitrary table joins. To address these issues, we propose a novel synthesis framework that incorporates key relationships from schema, imposes strong typing, and conducts schema-distance-weighted column sampling. We also adopt an intermediate representation (IR) for the SQL-to-text task to further improve the quality of the generated natural language questions. When existing powerful semantic parsers are pre-finetuned on our high-quality synthesized data, our experiments…

Multilingual BERT Post-Pretraining Alignment

Lin Pan, Chung-Wei Hang, Haode Qi +3 more

cs.CL · · cited 54× · arXiv ↗ · more like this ·

We propose a simple method to align multilingual contextual embeddings as a post-pretraining step for improved zero-shot cross-lingual transferability of the pretrained models. Using parallel data, our method aligns embeddings on the word level through the recently proposed Translation Language Modeling objective as well as on the sentence level via contrastive learning and random input shuffling. We also perform sentence-level code-switching with English when finetuning on downstream tasks. On XNLI, our best model (initialized from mBERT) improves over mBERT by 4.7% in the zero-shot setting and achieves comparable result to XLM for translate-train while using less than 18% of the same parallel data and 31% less model parameters. On MLQA, our model outperforms XLM-R_Base that has 57% more parameters than ours.

Dissecting multilingual Beijing: the space and scale of vernacular globalization

Lin Pan

Linguistics and Language · Visual Communication · · cited 53× · DOI ↗ · more like this ·

Following Scollon and Scollon’s Discourse in Place: Language in the Material World (2003) in which the authors proposed that the placement and lettering of signs in public places convey the sociopolitical meaning of the region and society, this article employs the geosemiotic principles that they introduced to explore how regimes of languages are organized in fast globalizing Beijing. In doing so, the author incorporates the ‘space and scale’ perspective initiated by Immanuel Wallerstein and enriched by Blommaert, in an attempt to anchor the analysis of the phenomenology of multilingual language usage in a social and ideological context. She examines how the occurrence of standard and non-standard multilingual patterns in the streets of Beijing, a fast globalizing city, reflects people’s unequal access to global linguistic resources and marks the social stratification and power…

Multilingual Transfer Learning for QA Using Translation as Data Augmentation

Mihaela Bornea, Lin Pan, Sara Rosenthal +2 more

cs.CL · · cited 46× · arXiv ↗ · more like this ·

Prior work on multilingual question answering has mostly focused on using large multilingual pre-trained language models (LM) to perform zero-shot language-wise learning: train a QA model on English and test on other languages. In this work, we explore strategies that improve cross-lingual transfer by bringing the multilingual embeddings closer in the semantic space. Our first strategy augments the original English training data with machine translation-generated data. This results in a corpus of multilingual silver-labeled QA pairs that is 14 times larger than the original training set. In addition, we propose two novel strategies, language adversarial training and language arbitration framework, which significantly improve the (zero-resource) cross-lingual transfer performance and result in LM embeddings that are less language-variant. Empirically, we show that the proposed models…

Is English a threat to Chinese language and culture?

Lin Pan, Philip Seargeant

Linguistics and Language · English Today · · cited 43× · DOI ↗ · more like this ·

In an article published in English Today in 2005, Niu and Woolf contend that: EFL is a modern day Trojan horse filled with EFL teachers/soldiers or missionaries, armed with English words rather than bullets, intent upon re-colonizing the world to remake it in the image of Western democracy. China has brought the Trojan Horse within its gates and the army of EFL teachers is hard at work Westernizing China.

Advance artificial time series forecasting model for oil production using neuro fuzzy-based slime mould algorithm

Ayman Mutahar AlRassas, Mohammed A. A. Al‐qaness, Ahmed A. Ewees +4 more

Artificial Intelligence · Journal of Petroleum Exploration and Production Technology · · cited 38× · DOI ↗ · more like this ·

Oil production forecasting is an important task to manage petroleum reservoirs operations. In this study, a developed time series forecasting model is proposed for oil production using a new improved version of the adaptive neuro-fuzzy inference system (ANFIS). This model is improved by using an optimization algorithm, the slime mould algorithm (SMA). The SMA is a new algorithm that is applied for solving different optimization tasks. However, its search mechanism suffers from some limitations, for example, trapping at local optima. Thus, we modify the SMA using an intelligence search technique called opposition-based learning (OLB). The developed model, ANFIS-SMAOLB, is evaluated with different real-world oil production data collected from two oilfields in two different countries, Masila oilfield (Yemen) and Tahe oilfield (China). Furthermore, the evaluation of this model is considered…

Revisiting Image-Restoration Strategies: An Integrated Case Study of Three Athlete Sex Scandals in Sports News

Juan Meng, Po‐Lin Pan

Communication · International Journal of Sport Communication · · cited 37× · DOI ↗ · more like this ·

A limited number of studies have examined the effectiveness of apology techniques in image restoration of athletes involved in sex scandals. This case study used Benoit’s (1995) image-restoration strategies to examine the apology techniques 3 athletes used to negotiate their sex scandals and attempted to encourage further discussion of these techniques. Three athletes’ sex scandals were comparatively analyzed, including those of golfer Tiger Woods, National Football League quarterback Ben Roethlisberger, and the National Basketball Association’s Kobe Bryant. This case study integrated the apology statements made by each athlete and examined sports-news coverage of the scandals from The New York Times, USA Today, and The Washington Post . This case study offers important insights on how these athletes restored their images and handled the crises surrounded their sex scandals.

Multi-Source Domain Adaptation for Medical Image Segmentation

Chenhao Pei, Fuping Wu, Mingjing Yang +5 more

Artificial Intelligence · IEEE Transactions on Medical Imaging · · cited 30× · DOI ↗ · more like this ·

Unsupervised domain adaptation(UDA) aims to mitigate the performance drop of models tested on the target domain, due to the domain shift from the target to sources. Most UDA segmentation methods focus on the scenario of solely single source domain. However, in practical situations data with gold standard could be available from multiple sources (domains), and the multi-source training data could provide more information for knowledge transfer. How to utilize them to achieve better domain adaptation yet remains to be further explored. This work investigates multi-source UDA and proposes a new framework for medical image segmentation. Firstly, we employ a multi-level adversarial learning scheme to adapt features at different levels between each of the source domains and the target, to improve the segmentation performance. Then, we propose a multi-model consistency loss to transfer the…

The TechQA Dataset

Vittorio Castelli, Rishav Chakravarti, Saswati Dana +18 more

cs.CL · · cited 25× · arXiv ↗ · more like this ·

We introduce TechQA, a domain-adaptation question answering dataset for the technical support domain. The TechQA corpus highlights two real-world issues from the automated customer support domain. First, it contains actual questions posed by users on a technical forum, rather than questions generated specifically for a competition or a task. Second, it has a real-world size -- 600 training, 310 dev, and 490 evaluation question/answer pairs -- thus reflecting the cost of creating large labeled datasets with actual data. Consequently, TechQA is meant to stimulate research in domain adaptation rather than being a resource to build QA systems from scratch. The dataset was obtained by crawling the IBM Developer and IBM DeveloperWorks forums for questions with accepted answers that appear in a published IBM Technote---a technical document that addresses a specific technical issue. We also…

Automated Testing of WS-BPEL Service Compositions: A Scenario-Oriented Approach

Chang‐ai Sun, Yan Zhao, Lin Pan +2 more

Information Systems · IEEE Transactions on Services Computing · · cited 22× · DOI ↗ · more like this ·

Nowadays, service oriented architecture (SOA) has become one mainstream paradigm for developing distributed applications. As the basic unit in SOA, web services can be composed to construct complex applications. The quality of web services and their compositions is critical to the success of SOA applications. Testing, as a major quality assurance technique, is confronted with new challenges in the context of service compositions. In this paper, we propose a scenario-oriented testing approach that can automatically generate test cases for service compositions. Our approach is particularly focused on the service compositions specified by Business Process Execution Language for web services (WS-BPEL), a widely recognized executable service composition language. In the approach, a WS-BPEL service composition is first abstracted into a graph model; test scenarios are then derived from the…

Research on the Construction of Age-Friendly Community Based on Fuzzy Comprehensive Evaluation Model: Evidence from Community in Hefei of China

Lin Pan, Jian Sun, Rongting Zhou

Health · Risk Management and Healthcare Policy · · cited 18× · DOI ↗ · more like this ·

BACKGROUND AND AIM: The construction of an age-friendly community is beneficial to meeting the needs of the elderly and coping with ageing. It is significant to evaluate the quality of age-friendly communities through scientific methods. The study aims to evaluate the livability of communities among the elderly in China. METHODS: The data used in this study were sourced from the community of Hefei in China, including 1426 participants aged 60 and above. Taking community in Hefei of China as an example, the analytic hierarchy process fuzzy comprehensive evaluation model was used to evaluate the community. Furthermore, this paper constructs an evaluation index system for the livable elderly community and evaluates the comprehensive community system with the fuzzy comprehensive evaluation method. RESULTS: This study constructs the evaluation index system of an age-friendly community that…

Identification of a dual TAOK1 and MAP4K5 inhibitor using a structure-based virtual screening approach

Min‐Wu Chao, Tony Eight Lin, Wei‐Chun HuangFu +9 more

Computational Theory and Mathematics · Journal of Enzyme Inhibition and Medicinal Chemistry · · cited 18× · DOI ↗ · more like this ·

The STE20 kinase family is a complex signalling cascade that regulates cytoskeletal organisation and modulates the stress response. This signalling cascade includes various kinase mediators, such as TAOK1 and MAP4K5. The dysregulation of the STE20 kinase pathway is linked with cancer malignancy. A small-molecule inhibitor targeting the STE20 kinase pathway has therapeutic potential. In this study, a structure-based virtual screening (SBVS) approach was used to identify potential dual TAOK1 and MAP4K5 inhibitors. Enzymatic assays confirmed three potential dual inhibitors (>50% inhibition) from our virtual screening, and analysis of the TAOK1 and MAP4K5 binding sites indicated common interactions for dual inhibition. Compound 1 revealed potent inhibition of colorectal and lung cancer cell lines. Furthermore, compound 1 arrested cancer cells in the G0/G1 phase, which suggests the induction…

Spatial Accessibility Assessment of Prehospital EMS with a Focus on the Elderly Population: A Case Study in Ningbo, China

Huanhuan Zhu, Lin Pan, Yiji Li +7 more

Transportation · International Journal of Environmental Research and Public Health · · cited 17× · DOI ↗ · more like this ·

The spatial accessibility of prehospital EMS is particularly important for the elderly population's physiological functions. Due to the recent expansion of aging populations all over the globe, elderly people's spatial accessibility to prehospital EMS presents a serious challenge. An efficient strategy to address this issue involves using geographic information systems (GIS)-based tools to evaluate the spatial accessibility in conjunction with the spatial distribution of aging people, available road networks, and prehospital EMS facilities. This study employed gravity model and empirical Bayesian Kriging (EBK) interpolation analysis to evaluate the elderly's spatial access to prehospital EMS in Ningbo, China. In our study, we aimed to solve the following specific research questions: In the study area, "what are the characteristics of the prehospital EMS demand of the elderly?" "Do the…

Optimized Random Vector Functional Link network to predict oil production from Tahe oil field in China

Ahmed Alalimi, Lin Pan, Mohammed A. A. Al‐qaness +3 more

Artificial Intelligence · Oil & Gas Science and Technology – Revue d’IFP Energies nouvelles · · cited 14× · DOI ↗ · more like this ·

In China, Tahe Triassic oil field block 9 reservoir was discovered in 2002 by drilling wells S95 and S100. The distribution of the reservoir sand body is not clear. Therefore, it is necessary to study and to predict oil production from this oil field. In this study, we propose an improved Random Vector Functional Link (RVFL) network to predict oil production from Tahe oil field in China. The Spherical Search Optimizer (SSO) is applied to optimize the RVFL and to enhance its performance, where SSO works as a local search method that improved the parameters of the RVFL. We used a historical dataset of this oil field from 2002 to 2014 collected by a local partner. Our proposed model, called SSO-RVFL, has been evaluated with extensive comparisons to several optimization methods. The outcomes showed that, SSO-RVFL achieved accurate predictions and the SSO outperformed several optimization…

Frustratingly Easy Natural Question Answering

Lin Pan, Rishav Chakravarti, Anthony Ferritto +5 more

cs.CL · · cited 14× · arXiv ↗ · more like this ·

Existing literature on Question Answering (QA) mostly focuses on algorithmic novelty, data augmentation, or increasingly large pre-trained language models like XLNet and RoBERTa. Additionally, a lot of systems on the QA leaderboards do not have associated research documentation in order to successfully replicate their experiments. In this paper, we outline these algorithmic components such as Attention-over-Attention, coupled with data augmentation and ensembling strategies that have shown to yield state-of-the-art results on benchmark datasets like SQuAD, even achieving super-human performance. Contrary to these prior results, when we evaluate on the recently proposed Natural Questions benchmark dataset, we find that an incredibly simple approach of transfer learning from BERT outperforms the previous state-of-the-art system trained on 4 million more examples than ours by 1.9 F1…

Task Scheduling and Resource Allocation of Cloud Computing Based on QoS

Bai Lin Pan, Yan Ping Wang, Han Xi Li +1 more

Information Systems · Advanced materials research · · cited 14× · DOI ↗ · more like this ·

With the enlargement of the scope of cloud computing application, the number of users and types also increases accordingly, the special demand for cloud computing resources has also improved. Cloud computing task scheduling and resource allocation are key technologies, mainly responsible for assigning user jobs to the appropriate resources to perform. But the existing scheduling algorithm is not fully consider the user demand for resources is different, and not well provided for different users to meet the requirements of its resources. As the demand for quality of service based on cloud computing and cloud computing original scheduling algorithm, the computing power scheduling algorithm is proposed based on the QoS constraints to research the cloud computing task scheduling and resource allocation problems, improving the overall efficiency of cloud computing system.

Automatic removal of eye-blink artifacts based on ICA and peak detection algorithm

Junfeng Gao, Lin Pan, Yong Yang +1 more

Signal Processing · · cited 14× · DOI ↗ · more like this ·

In this study a novel technique is proposed to automatically remove eye-blink artifacts from EEG signals. EEG signal is decomposed into independent components by independent component analysis (ICA) method. First, we demonstrate that feature-extraction method is not very suitable for identifying eye-blink artifacts components. Hence a peak detection algorithm of independent component is proposed to identify eye-blink artifacts components. Finally, the proposed artifact removal method is evaluated on some contaminated EEG signals, and the results show that it is able to effectively remove eye-blink artifacts from EEG signals with little distortion of underlying brain signals.

Sport team-endorsed brands on Facebook

Joe Phua, Po‐Lin Pan, Kuan‐Ju Chen

Gender Studies · Online Information Review · · cited 13× · DOI ↗ · more like this ·

Purpose Applying social identity theory, the social identity-brand equity model and excitation-transfer theory, the purpose of this paper is to examine effects of game outcome (win/loss) and location (home/away) on sport fans’ brand attitude and purchase intention toward a brand endorsed by their favorite sport team on Facebook, as well as the mediating role of team identification. Design/methodology/approach A two (win/loss) by two (home/away) full-factorial between-subjects experiment was conducted during the US National Collegiate Athletic Association (NCAA) football season over a four-month period. Participants ( n =338), who were Facebook users and fans of a NCAA division I football team, completed an online questionnaire assessing brand attitude and purchase intention toward a team-endorsed brand on Facebook, during weeks after the team: won a home game, lost a home game, won an…

Coping with the milk scandal

Li Zeng, Lijie Zhou, Po‐Lin Pan +1 more

Communication · Journal of Communication Management · · cited 12× · DOI ↗ · more like this ·

Purpose The purpose of this paper is to examine crisis communication strategies used by four leading Chinese milk companies at various crisis stages to cope with the largest food safety crisis in China. Approaching the interplay between the situational crisis communication theory (Coombs, 2007) and the image restoration theory (Benoit, 1995), the study attempted to understand the importance of crisis management at various crisis stages by comparing crisis communication strategies used by surviving and thriving companies with those by the company that failed. Design/methodology/approach Dividing a crisis management process into three stages, a content analysis was conducted to analyze five major crisis communication strategies – evasion of responsibility, rebuilding, bolstering, endorsement of outside experts and government relations – used by Chinese milk companies at various crisis…