Search

Online deception is disrupting our daily life, organizational process, and even national security. Existing approaches to online deception detection follow a traditional paradigm by using a set of cues as antecedents for deception detection, which may be hindered by ineffective cue identification. Motivated by the strength of statistical language models (SLMs) in capturing the dependency of words in text without explicit feature extraction, we developed SLMs to detect online deception. We also addressed the data-sparsity problem in building SLMs in general and in deception detection in specific using smoothing and vocabulary pruning techniques. The developed SLMs were evaluated empirically with diverse data sets. The results showed that the proposed SLM approach to deception detection outperformed a state-of-the-art text categorization method, as well as traditional feature-based methods. [ABSTRACT FROM AUTHOR]

Recent advances in Artificial Intelligence (AI), such as the development of large language models like ChatGPT, have blurred the boundaries between human and AI-generated text. This has led to a pressing need for tools that can determine whether text has been created or revised using AI. A general and universally effective detection model would be extremely useful, but appears to be beyond the reach of current technology and detection methods. The research described in this study adopts a domain and task specific approach and shows that specialized detection models can attain high accuracy. The study focuses on the higher education graduate admissions process, with the specific goal of identifying AI-generated and AI-revised Letters of Recommendation (LORs) and Statements of Intent (SOIs). Detecting such application materials is essential to ensure that applicants are evaluated on their true merits and abilities, and to foster an equitable and trustworthy admissions process. Our research is based on 3755 LORs and 1973 SOIs extracted from the application records of Fordham University's Master's programs in Computer Science and Data Science. To facilitate the construction and evaluation of detection models, we generated AI counterparts for each LOR and SOI using the GPT-3.5 Turbo API. The prompts for AI-generation text were derived from the admission data of the respective applicants, and the AI-revised LORs and SOIs were generated directly from the human-authored versions. We also utilize an open-access GPT-wiki-intro dataset to further validate our hypothesis regarding the feasibility of constructing domain-specific AI content detectors. Our experiments yield promising results in developing classifiers tailored to a specific domain when provided with sufficient training samples. Additionally, we present a comparative analysis of the word frequency and statistical characteristics of the text, which provides convincing evidence that ChatGPT employs distinctive vocabulary and paragraph structure compared to human-authored text. The code for this study is available on GitHub, and the models can be executed on user-provided data via an interactive web interface. [ABSTRACT FROM AUTHOR]

Log messages from enterprise-level software systems contain crucial runtime details. Engineers can convert log messages into structured data through log parsing, laying the foundation for downstream tasks such as log anomaly detection. Existing log parsing schemes usually underperform in production environments for several reasons: first, they often ignore the semantics of log messages; second, they are often not adapted to different systems, and their performance varies greatly; and finally, they are difficult to adapt to the complexity and variety of log formats in the real environment. In response to the limitations of current approaches, we introduce IPLog (Intelligent Parse Log), a parsing method designed to address these issues. IPLog samples a limited set of log samples based on the distribution of templates in the system's historical logs, and allows the model to make full use of the small number of log samples to recognize common patterns of keywords and parameters through few-shot learning, and thus can be easily adapted to different systems. In addition, IPLog can further improve the grouping accuracy of log templates through a novel manual feedback merge query strategy based on the longest common prefix, thus enhancing the model's adaptability to handle complex log formats in production environments. We conducted experiments on four newly released public log datasets, and the experimental results show that IPLog can achieve an average grouping accuracy (GA) of 0.987 and parsing accuracy (PA) of 0.914 on the four public datasets, which are the best among the mainstream parsing schemes. These results demonstrate that IPLog is effective for log parsing tasks. [ABSTRACT FROM AUTHOR]

Social media platforms have completely revolutionized human communication and social interactions. Their positive impacts are simply undeniable. What has also become undeniable is the prevalence of harmful antisocial behaviors on these platforms. Cyberbullying, misinformation, hate speech, radicalization, and extremist propaganda have caused significant harms to society and its most vulnerable populations. Thus, the social media forensics field was born to enable investigators and law enforcement agents to better investigate and prosecute these cybercrimes. This paper surveys the latest research works in the field to explore how artificial intelligence (AI) techniques are being utilized in social media forensics investigations. We examine how natural language processing can be used to identify extremist ideologies, detect online bullying, and analyze deceptive profiles. Additionally, we explore the literature on GNNs and how they are applied in social network modeling for forensic purposes. We conclude by discussing the key challenges in the field and suggest future research directions. [ABSTRACT FROM AUTHOR]

We present the results of an ongoing collaboration between computer science and psychology researchers that employs Natural Language Processing (NLP) methods to examine the trajectory of semantic space used during group idea generation sessions. Specifically, we track and estimate the region of semantic space being used and the degree to which new ideas expand that space. We present a visualization of this space mapping endeavor and compare human ratings of creativity dimensions (i.e., novelty, task-relevance, and elaboration) to algorithm-based estimations of those same dimensions. The semantic space mapping and algorithm development projects can be used to deliver real-time feedback to human creative groups in order to optimize the collaborative creativity process. The overall goal of this research is to increase the "survival" of novel ideas and their elaboration in the collaborative ideation and subsequent decision processes. [ABSTRACT FROM AUTHOR]

This article shows that the collaboration between social science and computer science scholars proves fruitful in enhancing conceptual and methodological innovation in research appropriate for the digital world. It presents arguments for ways in which a multi-disciplinary approach can strengthen media studies and innovatively advance both research breadth and depth. To illustrate this interesting connection of both disciplines, we present the example analysis of large data from Twitter and discuss this analysis in a communication science research environment. We propose TwiNeR, a software tool that analyzes tweet content using an advanced language modeling approach for classifying tweets into five prototypical messages referring to 'activities' related to news and news sources in the Twitter network (i.e., source-fed article, userfed article, content spread by user, other source content, other user content). [ABSTRACT FROM AUTHOR]

This essay is a brief introduction to salient issues for higher education surrounding the emergence of ChatGPT. Issues include the tension in distinguishing knowledge vs. information, understanding fundamental differences between human languages and ChatGPT, the reimagining and recasting of assessment and testing, potential impacts on cognitive and linguistic proficiencies, potential intensification of existing inequalities in access to technologies for fragile and disadvantaged communities, and proliferation of implicit biases. [ABSTRACT FROM AUTHOR]

There has been increased attention toward the datasets that are used to train and build AI technologies from the computer science and social science research communities, but less from legal scholarship. Both Large-Scale Language Datasets (LSLDs) and Large-Scale Computer Vision Datasets (LSCVDs) have been at the forefront of such discussions, due to recent controversies involving the use of facial recognition technologies, and the discussion of the use of publiclyavailable textfor the training of massive models which generate humanlike text. Many of these datasets serve as "benchmarks" to develop models that are used both in academic and industry research, while others are used solely for training models. The process of developing LSLDs and LSCVDs is complex and contextual, involving dozens of decisions about what kinds of data to collect, label, and train a model on, as well as how to make the data available to other researchers. However, little attention has been paid to mapping and consolidating the legal issues that arise at different stages of this process: when the data is being collected, after the data is used to build and evaluate models and applications, and how that data is distributed more widely. In this article, we offer four main contributions. First, we describe what kinds of objects these datasets are, how many different kinds exist, what types of modalities they encompass, and why they are important. Second, we provide more clarity about the stages of dataset development - a process that has thus far been subsumed within broader discussions about bias and discrimination - and the subjects who may be susceptible to harms at each point of development. Third, we provide a matrix of both the stages of dataset development and the subjects of dataset development, which traces the connections between stages and subjects. Fourth, we use this analysis to identify some basic legal issues that arise at the various stages in order to foster a better understanding of the dilemmas and tensions that arise at every stage. We situate our discussion within wider discussion of current debates and proposals related to algorithmic accountability. This paper fulfills an essential gap when it comes to comprehending the complicated landscape of legal issues connected to datasets and the gigantic AI models trained on them. [ABSTRACT FROM AUTHOR]

There has been considerable work recently in the natural language community and elsewhere on Responsible AI. Much of this work focuses on fairness and biases (henceforth Risks 1.0), following the 2016 best seller: Weapons of Math Destruction. Two books published in 2022,  The Chaos Machine and Like, Comment, Subscribe , raise additional risks to public health/safety/security such as genocide, insurrection, polarized politics, vaccinations (henceforth, Risks 2.0). These books suggest that the use of machine learning to maximize engagement in social media has created a Frankenstein Monster that is exploiting human weaknesses with persuasive technology, the illusory truth effect, Pavlovian conditioning, and Skinner's intermittent variable reinforcement. Just as we cannot expect tobacco companies to sell fewer cigarettes and prioritize public health ahead of profits, so too, it may be asking too much of companies (and countries) to stop trafficking in misinformation given that it is so effective and so insanely profitable (at least in the short term). Eventually, we believe the current chaos will end, like the lawlessness in Wild West, because chaos is bad for business. As computer scientists, this paper will summarize criticisms from other fields and focus on implications for computer science; we will not attempt to contribute to those other fields. There is quite a bit of work in computer science on these risks, especially on Risks 1.0 (bias and fairness), but more work is needed, especially on Risks 2.0 (addictive, dangerous, and deadly). [ABSTRACT FROM AUTHOR]

Identifying significant shots in a rally is important for evaluating players' performance in badminton matches. While there are several studies that have quantified player performance in other sports, analyzing badminton data has remained untouched. In this article, we introduce a badminton language to fully describe the process of the shot, and we propose a deep-learning model composed of a novel short-term extractor and a long-term encoder for capturing a shot-by-shot sequence in a badminton rally by framing the problem as predicting a rally result. Our model incorporates an attention mechanism to enable the transparency between the action sequence and the rally result, which is essential for badminton experts to gain interpretable predictions. Experimental evaluation based on a real-world dataset demonstrates that our proposed model outperforms the strong baselines. We also conducted case studies to show the ability to enhance players' decision-making confidence and to provide advanced insights for coaching, which benefits the badminton analysis community and bridges the gap between the field of badminton and computer science. [ABSTRACT FROM AUTHOR]

Peer review is a key component of the publishing process in most fields of science. Increasing submission rates put a strain on reviewing quality and efficiency, motivating the development of applications to support the reviewing and editorial work. While existing NLP studies focus on the analysis of individual texts, editorial assistance often requires modeling interactions between pairs of texts—yet general frameworks and datasets to support this scenario are missing. Relationships between texts are the core object of the intertextuality theory—a family of approaches in literary studies not yet operationalized in NLP. Inspired by prior theoretical work, we propose the first intertextual model of text-based collaboration, which encompasses three major phenomena that make up a full iteration of the review–revise–and–resubmit cycle: pragmatic tagging, linking, and long-document version alignment. While peer review is used across the fields of science and publication formats, existing datasets solely focus on conference-style review in computer science. Addressing this, we instantiate our proposed model in the first annotated multidomain corpus in journal-style post-publication open peer review, and provide detailed insights into the practical aspects of intertextual annotation. Our resource is a major step toward multidomain, fine-grained applications of NLP in editorial support for peer review, and our intertextual framework paves the path for general-purpose modeling of text-based collaboration. We make our corpus, detailed annotation guidelines, and accompanying code publicly available. [ABSTRACT FROM AUTHOR]

Textbook Question Answering (TQA) is the task of answering diagram and non-diagram questions given large multi-modal contexts consisting of abundant text and diagrams. Deep text understandings and effective learning of diagram semantics are important for this task due to its specificity. In this paper, we propose a Weakly Supervised learning method for TQA (WSTQ), which regards the incompletely accurate results of essential intermediate procedures for this task as supervision to develop Text Matching (TM) and Relation Detection (RD) tasks and then employs the tasks to motivate itself to learn strong text comprehension and excellent diagram semantics respectively. Specifically, we apply the result of text retrieval to build positive as well as negative text pairs. In order to learn deep text understandings, we first pre-train the text understanding module of WSTQ on TM and then fine-tune it on TQA. We build positive as well as negative relation pairs by checking whether there is any overlap between the items/regions detected from diagrams using object detection. The RD task forces our method to learn the relationships between regions, which are crucial to express the diagram semantics. We train WSTQ on RD and TQA simultaneously, i.e., multitask learning, to obtain effective diagram semantics and then improve the TQA performance. Extensive experiments are carried out on CK12-QA and AI2D to verify the effectiveness of WSTQ. Experimental results show that our method achieves significant accuracy improvements of 5.02% and 4.12% on test splits of the above datasets respectively than the current state-of-the-art baseline. We have released our code on https://github.com/dr-majie/WSTQ. [ABSTRACT FROM AUTHOR]

Sender-receiver interactions, and specifically persuasion games, are widely researched in economic modeling and artificial intelligence, and serve as a solid foundation for powerful applications. However, in the classic persuasion games setting, the messages sent from the expert to the decision-maker are abstract or well-structured application-specific signals rather than natural (human) language messages, although natural language is a very common communication signal in real-world persuasion setups. This paper addresses the use of natural language in persuasion games, exploring its impact on the decisions made by the players and aiming to construct effective models for the prediction of these decisions. For this purpose, we conduct an online repeated interaction experiment. At each trial of the interaction, an informed expert aims to sell an uninformed decision-maker a vacation in a hotel, by sending her a review that describes the hotel. While the expert is exposed to several scored reviews, the decision-maker observes only the single review sent by the expert, and her payoff in case she chooses to take the hotel is a random draw from the review score distribution available to the expert only. The expert’s payoff, in turn, depends on the number of times the decision-maker chooses the hotel. We also compare the behavioral patterns in this experiment to the equivalent patterns in similar experiments where the communication is based on the numerical values of the reviews rather than the reviews’ text, and observe substantial differences which can be explained through an equilibrium analysis of the game. We consider a number of modeling approaches for our verbal communication setup, differing from each other in the model type (deep neural network (DNN) vs. linear classifier), the type of features used by the model (textual, behavioral or both) and the source of the textual features (DNN-based vs. hand-crafted). Our results demonstrate that given a prefix of the interaction sequence, our models can predict the future decisions of the decision-maker, particularly when a sequential modeling approach and hand-crafted textual features are applied. Further analysis of the hand-crafted textual features allows us to make initial observations about the aspects of text that drive decision making in our setup.1 [ABSTRACT FROM AUTHOR]

Developing artificial learning systems that can understand and generate natural language has been one of the long-standing goals of artificial intelligence. Recent decades have witnessed an impressive progress on both of these problems, giving rise to a new family of approaches. Especially, the advances in deep learning over the past couple of years have led to neural approaches to natural language generation (NLG). These methods combine generative language learning techniques with neural-networks based frameworks. With a wide range of applications in natural language processing, neural NLG (NNLG) is a new and fast growing field of research. In this state-of-the-art report, we investigate the recent developments and applications of NNLG in its full extent from a multidimensional view, covering critical perspectives such as multimodality, multilinguality, controllability and learning strategies. We summarize the fundamental building blocks of NNLG approaches from these aspects and provide detailed reviews of commonly used preprocessing steps and basic neural architectures. This report also focuses on the seminal applications of these NNLG models such as machine translation, description generation, automatic speech recognition, abstractive summarization, text simplification, question answering and generation, and dialogue generation. Finally, we conclude with a thorough discussion of the described frameworks by pointing out some open research directions. [ABSTRACT FROM AUTHOR]

Biomedical terms extracted using Word2vec, the most popular word embedding model in recent years, serve as the foundation for various natural language processing (NLP) applications, such as biomedical information retrieval, relation extraction, and recommendation systems. The objective of this study is to examine how changes in the ratio of the biomedical domain to general domain data in the corpus affect the extraction of similar biomedical terms using Word2vec. We downloaded abstracts of 214,892 articles from PubMed Central (PMC) and the 3.9 GB Billion Word (BW) benchmark corpus from the computer science community. The datasets were preprocessed and grouped into 11 corpora based on the ratio of BW to PMC, ranging from 0:10 to 10:0, and then Word2vec models were trained on these corpora. The cosine similarities between the biomedical terms obtained from the Word2vec models were then compared in each model. The results indicated that the models trained with both BW and PMC data outperformed the model trained only with medical data. The similarity between the biomedical terms extracted by the Word2vec model increased when the ratio of the biomedical domain to general domain data was 3:7 to 5:5. This study allows NLP researchers to apply Word2vec based on more information and increase the similarity of extracted biomedical terms to improve their effectiveness in NLP applications, such as biomedical information extraction. [ABSTRACT FROM AUTHOR]

Code-switching (CS) is prevalent in bilingual communication. With the bibliometric method CiteSpace, this paper gives an overview of the 1319 bibliometric records (1998–2018) on CS selected from the Web of Science. A co-occurring keyword network and a co-citation network of these publications were constructed. Analyses on the high-frequency keywords, document co-citation, co-cited reference clusters, publications with centrality betweenness and citation bursts have been conducted. The results indicate that the landscape of CS research has been expanded from the initial focus on the lexical/conceptual aspect to today's in-depth explorations on language contact, language control, systematic pedagogy, etc. Topics related to 'multimodality', 'translanguaging', 'switch costs' and 'automatic speech recognition' can represent the emerging trends of CS studies in various perspectives, such as linguistics, social science, cognitive science and computer science. [ABSTRACT FROM AUTHOR]

The recent years have witnessed a rapid increase in the number of scientific articles in biomedical domain. These literature are mostly available and readily accessible in electronic format. The domain knowledge hidden in them is critical for biomedical research and applications, which makes biomedical literature mining (BLM) techniques highly demanding. Numerous efforts have been made on this topic from both biomedical informatics (BMI) and computer science (CS) communities. The BMI community focuses more on the concrete application problems and thus prefer more interpretable and descriptive methods, while the CS community chases more on superior performance and generalization ability, thus more sophisticated and universal models are developed. The goal of this paper is to provide a review of the recent advances in BLM from both communities and inspire new research directions. [ABSTRACT FROM AUTHOR]

Developing an intelligent chatbot has evolved in the last few years to become a trending topic in the area of computer science. However, a chatbot often fails to understand the user's intent, which can lead to the generation of inappropriate responses that cause dialogue breakdown and user dissatisfaction. Detecting the dialogue breakdown is essential to improve the performance of the chatbot and increase user satisfaction. Recent approaches have focused on modeling conversation breakdown using serveral approaches, including supervised and unsupervised approaches. Unsupervised approach relay heavy datasets, which make it challenging to apply it to the breakdown task. Another challenge facing predicting breakdown in conversation is the bias of human annotation for the dataset and the handling process for the breakdown. To tackle this challenge, we have developed a supervised ensemble automated approach that measures Chatbot Quality of Service (CQoS) based on dialogue breakdown. The proposed approach is able to label the datasets based on sentiment considering the context of the conversion to predict the breakdown. In this paper we aim to detect the affect of sentiment change of each speaker in a conversation. Furthermore, we use the supervised ensemble model to measure the CQoS based on breakdown. Then we handle this problem by using a hand-over mechanism that transfers the user to a live agent. Based on this idea, we perform several experiments across several datasets and state-of-the-art models, and we find that using sentiment as a trigger for breakdown outperforms human annotation. Overall, we infer that knowledge acquired from the supervised ensemble model can indeed help to measure CQoS based on detecting the breakdown in conversation. [ABSTRACT FROM AUTHOR]

We understand this paper as a contribution to the "anatomy" of conceptual models. We propose a signature of conceptual models for their characterization, which allows a clear distinction from other types of models. The motivation for this work arose from the observation that conceptual models are widely discussed in science and practice, especially in computer science, but that their potential is far from being exploited. We combine our proposal of a more transparent explanation of the nature of conceptual models with an approach that classifies conceptual models as a link between the dimension of linguistic terms and the encyclopedic dimension of notions. As a paradigm we use the triptych, whose central tableau represents the model dimension. The effectiveness of this explanatory approach is illustrated by a number of examples. We derive a number of open research questions that should be answered to complete the anatomy of conceptual models. [ABSTRACT FROM AUTHOR]

A commonly used strategy to improve search accuracy is through feedback techniques. Most existing work on feedback relies on positive information, and has been extensively studied in information retrieval. However, when a query topic is difficult and the results from the first-pass retrieval are very poor, it is impossible to extract enough useful terms from a few positive documents. Therefore, the positive feedback strategy is incapable to improve retrieval in this situation. Contrarily, there is a relatively large number of negative documents in the top of the result list, and it has been confirmed that negative feedback strategy is an important and useful way for adapting this scenario by several recent studies. In this paper, we consider a scenario when the search results are so poor that there are at most three relevant documents in the top twenty documents. Then, we conduct a novel study of multiple strategies for relevance feedback using both positive and negative examples from the first-pass retrieval to improve retrieval accuracy for such difficult queries. Experimental results on these TREC collections show that the proposed language model based multiple model feedback method which is generally more effective than both the baseline method and the methods using only positive or negative model. [ABSTRACT FROM AUTHOR]

Neuroevolution is the name given to a field of computer science that applies evolutionary computation for evolving some aspects of neural networks. After the AI Winter came to an end, neural networks reemerged to solve a great variety of problems. However, their usage requires designing their topology, a decision with a potentially high impact on performance. Whereas many works have tried to suggest rules-of-thumb for designing topologies, the truth is that there are not analytic procedures for determining the optimal one for a given problem, and trial-and-error is often used instead. Neuroevolution arose almost 3 decades ago, with some works focusing on the evolutionary design of the topology and most works describing techniques for learning connection weights. Since then, evolutionary computation has been proved to be a convenient approach for determining the topology and weights of neural networks, and neuroevolution has been applied to a great variety of fields. However, for more than 2 decades neuroevolution has mainly focused on simple artificial neural networks models, far from today's deep learning standards. This is insufficient for determining good architectures for modern networks extensively used nowadays, which involve multiple hidden layers, recurrent cells, etc. More importantly, deep and convolutional neural networks have become a de facto standard in representation learning for solving many different problems, and neuroevolution has only focused in this kind of networks in very recent years, with many works being presented in 2017 onward. In this paper, we review the field of neuroevolution during the last 3 decades. We will put the focus on very recent works on the evolution of deep and convolutional neural networks, which is a new but growing field of study. To the best of our knowledge, this is the best survey reviewing the literature in this field, and we have described the features of each work as well as their performance on well-known databases when available. This work aims to provide a complete reference of all works related to neuroevolution of convolutional neural networks up to the date. Finally, we will provide some future directions for the advancement of this research area. [ABSTRACT FROM AUTHOR]

We present a novel approach based on deep learning for decoding sensory information from non-invasively recorded Electroencephalograms (EEG). It can either be used in a passive Brain-Computer Interface (BCI) to predict properties of a visual stimulus the person is viewing, or it can be used to actively control a BCI application. Both scenarios were tested, whereby an average information transfer rate (ITR) of 701 bit/min was achieved for the passive BCI approach with the best subject achieving an online ITR of 1237 bit/min. Further, it allowed the discrimination of 500,000 different visual stimuli based on only 2 seconds of EEG data with an accuracy of up to 100%. When using the method for an asynchronous self-paced BCI for spelling, an average utility rate of 175 bit/min was achieved, which corresponds to an average of 35 error-free letters per minute. As the presented method extracts more than three times more information than the previously fastest approach, we suggest that EEG signals carry more information than generally assumed. Finally, we observed a ceiling effect such that information content in the EEG exceeds that required for BCI control, and therefore we discuss if BCI research has reached a point where the performance of non-invasive visual BCI control cannot be substantially improved anymore. [ABSTRACT FROM AUTHOR]

Network representation learning (NRL) aims to learn low-dimensional vectors for vertices in a network. Most existing NRL methods focus on learning representations from local context of vertices (such as their neighbors). Nevertheless, vertices in many complex networks also exhibit significant global patterns widely known as communities. It's intuitive that vertices in the same community tend to connect densely and share common attributes. These patterns are expected to improve NRL and benefit relevant evaluation tasks, such as link prediction and vertex classification. Inspired by the analogy between network representation learning and text modeling, we propose a unified NRL framework by introducing community information of vertices, named as Community-enhanced Network Representation Learning (CNRL). CNRL simultaneously detects community distribution of each vertex and learns embeddings of both vertices and communities. Moreover, the proposed community enhancement mechanism can be applied to various existing NRL models. In experiments, we evaluate our model on vertex classification, link prediction, and community detection using several real-world datasets. The results demonstrate that CNRL significantly and consistently outperforms other state-of-the-art methods while verifying our assumptions on the correlations between vertices and communities. [ABSTRACT FROM AUTHOR]