2021 IEEE 20th International Conference on Cognitive Informatics & Cognitive Computing (ICCI*CC)最新文献

{"title":"The Failure of Deep Neural Networks to Capture Human Language’s Cognitive Core","authors":"R. Berwick","doi":"10.1109/iccicc53683.2021.9811297","DOIUrl":"https://doi.org/10.1109/iccicc53683.2021.9811297","url":null,"abstract":"Current deep neural networks have made remarkable advances in their ability to analyze and use natural language, with great apparent engineering success. But how well do these systems mirror the cognitive constraints associated with human language? In this talk we show that there are three essential core computations that characterize human language as an engine of human thought. One is \"digital infinity\"– the fact that we can produce an open-ended countably infinite number of sentences. The second is that sentences are hierarchically structured, rather than being arranged in a linear array. The third property is that human language computations always admit the possibility of \"displacement\" – a word or phrase can be pronounced at a place distinct from its usual location of semantic interpretation. All three properties can be shown to follow from a single, simple, recursive combinatorial operation. We provide empirical evidence for all three properties, both from concrete developmental examples as well as psycholinguistic and brain imaging experiments.What about current \"deep neural network\" systems? Although they perform very well after large-scale training, their success appears to be grounded on accurate table-lookup–memorization–without truly mirroring the three key computational principles of human language cognition. By \"stress testing\" currently available deep neural network processors, we show that they are, perhaps surprisingly very fragile when presented even with simple examples that deviate modestly from the examples on which they were trained. In particular, they fail to properly represent hierarchical structure and they cannot reliably reconstruct examples of sentences with \"displacement\" if the examples go just a bit beyond the complexity of their training set data. For example, while a deep neural network system might work on \"Which cookie did Bob want,\" it fails on, \"Which cookie did Bob want to eat.\" Such failures indicate that the neural net systems have not generalized in the same sense that children do, since children can easily handle such examples after receiving much more limited training data.","PeriodicalId":101653,"journal":{"name":"2021 IEEE 20th International Conference on Cognitive Informatics & Cognitive Computing (ICCI*CC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115940326","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"IEEE ICCI*CC Series in Year 20: Latest Advances in Cognitive Computing (Plenary Panel Report-II)","authors":"W. Kinsner, Haibin Zhu, G. Baciu, G. Luo, Fernando Rubio, Jie Sui, Runhe Huang, H. Hiraishi, Jun Peng, Liang Chen","doi":"10.1109/ICCICC53683.2021.9811336","DOIUrl":"https://doi.org/10.1109/ICCICC53683.2021.9811336","url":null,"abstract":"Cognitive Computing (CC) is a contemporary field of fundamental intelligence theories and general AI technologies triggered by the transdisciplinary development in intelligence, computer, brain, knowledge, cognitive, robotic, and cybernetic sciences for engineering implementations. This paper presents a summary of the plenary panel (Part II) on the theoretical foundations of CI/CC and recent breakthroughs in AI engineering reported in the 20th IEEE International ICCI*CC Conference (ICCI*CC'21). The latest advances in CI and CC towards general AI are presented by twenty-two distinguished panelists. Strategic AI engineering applications in CI, CC, and cognitive systems are elaborated for abstract intelligence, cognitive robots, autonomous systems, intelligent vehicles, and safety-and-mission-critical systems.","PeriodicalId":101653,"journal":{"name":"2021 IEEE 20th International Conference on Cognitive Informatics & Cognitive Computing (ICCI*CC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114228175","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Autonomous Software Requirement Specifications towards AI Programming","authors":"Guoyin Wang, James Y. Xu","doi":"10.1109/ICCICC53683.2021.9811311","DOIUrl":"https://doi.org/10.1109/ICCICC53683.2021.9811311","url":null,"abstract":"Autonomous software requirement specifications and code generation are not only an ultimate goal of AI Programming (AIP), but also a persistent challenge to theories and technologies of software engineering. A cognitive system is demanded to autonomously elicit and rigorously refine software requirements in order to generate a set of formal specifications as the front-end of AIP. This paper presents a novel methodology for the design of an Intelligent Tool for Autonomous Software Specifications (ITASS) based on latest advances in software science and intelligent mathematics. ITASS is implemented as an interactive system for capturing software requirements and generating mathematic-based specifications for code generation in the back-end of the AIP system. The ITASS methodology and experiments are demonstrated for solving real-world and complex software engineering problems enabled by the AIP theories underpinned by intelligent mathematics.","PeriodicalId":101653,"journal":{"name":"2021 IEEE 20th International Conference on Cognitive Informatics & Cognitive Computing (ICCI*CC)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127346542","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Small and Medium-sized Enterprises Credit Risk Assessment Based on Temporal Knowledge Graphs","authors":"Chuanyang Hong, Mengyuan Tan, Siyu Wang, Junliang Wang, Mu Li, Jiangtao Qiu","doi":"10.1109/ICCICC53683.2021.9811323","DOIUrl":"https://doi.org/10.1109/ICCICC53683.2021.9811323","url":null,"abstract":"Credit Risk Assessment (CRA) is a challenging task in the financial field. Previous studies mainly focus on large firms with more comprehensive data especially financial data, annual reports, but for Small and Medium-sized Enterprises (SMEs), there is only public data to utilize, such as news, cases, etc. To better assess risk for SMEs, we constructed a temporal knowledge graph by using public data and proposed a credit risk assessment model (short for TKG-CRA) which comprehensively considers the topological structure of the temporal enterprise knowledge graph with the spread of risks and the neighbor node sequence. Experiments on real-world datasets prove that our model has a larger performance improvement than other traditional methods.","PeriodicalId":101653,"journal":{"name":"2021 IEEE 20th International Conference on Cognitive Informatics & Cognitive Computing (ICCI*CC)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123501767","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Interactive Visualization of Deep Learning for 3D Brain Data Analysis","authors":"Huang Li, S. Fang, J. Goñi, A. Saykin, Li Shen","doi":"10.1109/ICCICC53683.2021.9811312","DOIUrl":"https://doi.org/10.1109/ICCICC53683.2021.9811312","url":null,"abstract":"With multiple hidden layers and massive combinations of features and weights, deep learning models are hard to understand, and even more difficult to interact with. In this paper we describe a visual analytics platform to help with the understanding of and interaction with the deep learning process of human brain image data. A brain connectome network dataset is used to train a classifier for the diagnosis of Alzheimer's Disease (AD). 3D rendering of brain images is integrated into the interactive visualization process of a deep neural network to bring contextual information of the application to the analysis framework. A backpropagation algorithm is applied to track the image features that are captured by each node in the hidden layers. Our results demonstrate that interactive visualization can not only help the understanding of the deep learning process, but also provide a platform for domain experts to interact with and assist in the learning process, which can potentially enhance the interpretability and accuracy of the analysis.","PeriodicalId":101653,"journal":{"name":"2021 IEEE 20th International Conference on Cognitive Informatics & Cognitive Computing (ICCI*CC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131392947","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Designing a Neural Network and a Genetic- Algorithm-Based Adaptive Wavelet for Internet Traffic Containing DDoS Attacks","authors":"M. Ghanbari, W. Kinsner","doi":"10.1109/iccicc53683.2021.9811314","DOIUrl":"https://doi.org/10.1109/iccicc53683.2021.9811314","url":null,"abstract":"This paper presents the design of an adaptive mother wavelet for detecting Internet traffic data (ITD) with distributed denial of service (DDoS) attacks (DDoS ITD). The proposed procedure consists of designing an adaptive mother wavelet genetic neural network (GNN) for detecting the DDoS ITD, A multi-objective optimization based on a genetic algorithm is used to create a set of adaptive mother wavelets that best fit the weight parameters for a given input data recording. Moreover, a weighted cost function is used to measure how well the GNN is able to create a mother wavelet. The best mother wavelet coefficients for detecting DDoS attacks are achieved with coefficients [−0.3744,0.0034]. The created mother wavelet increased the detection rate of the DDoS attacks by 0.3% when compared to the Haar mother wavelet.","PeriodicalId":101653,"journal":{"name":"2021 IEEE 20th International Conference on Cognitive Informatics & Cognitive Computing (ICCI*CC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130328492","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A contract among autonomous agents to deal with egalitarian social welfare","authors":"Jonathan Carrero, Ismael Rodríguez, F. Rubio","doi":"10.1109/ICCICC53683.2021.9811317","DOIUrl":"https://doi.org/10.1109/ICCICC53683.2021.9811317","url":null,"abstract":"Auction security has been a major challenge for researchers in this area. For example, one of the biggest problems has always been the trust in a third party, an intermediary, which is the one who usually conducts the auction and knows the bids made by the participants. Over time, traditional methods have been overtaken by new technologies that eliminate the problems that arise when using traditional methods. Blockchain technology allows us to use its inherent characteristics of privacy, traceability and decentralization to conduct auctions with a much higher level of security and to execute auctions while reducing transaction costs. In addition, the automation of operations provided by smart contracts allows us to eliminate the intermediary, leading to additional cost savings. Furthermore, in contrast to previous technologies, the pseudo-anonymity of blockchain allows us to verify the authenticity of data, mitigating malicious behavior on the part of agents. In this paper we address this challenge; we present a smart contract that allows us to run an auction within the Ethereum blockchain at a relatively low cost, eliminating the intermediary and guaranteeing the trust of the agents involved in the auction. In particular, we concentrate on dealing with egalitarian social welfare, where the goal is to maximize the utility of the agent whose utility turns out to be minimal.","PeriodicalId":101653,"journal":{"name":"2021 IEEE 20th International Conference on Cognitive Informatics & Cognitive Computing (ICCI*CC)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128951222","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Individual identification model and method for estimating social rank among herd of dairy cows using YOLOv5","authors":"Tom Uchino, H. Ohwada","doi":"10.1109/ICCICC53683.2021.9811319","DOIUrl":"https://doi.org/10.1109/ICCICC53683.2021.9811319","url":null,"abstract":"Animals typically have a hierarchical relationship called social rank. For dairy cows, this rank is particularly important in agriculture because it affects milk production, disease, and the accuracy of estrus detection. The social rank of dairy cows has been studied for a long time; it is determined manually by monitoring the behavior of dairy cows, and this requires a significant amount of time and experience. Thus, in this study, a method for automatically estimating the social rank of a herd using video images obtained from cameras is proposed. In particular, the method can automatically determine the social ranking of all cows by extracting and analyzing the fighting behavior depicted in the video images.Specifically, we used YOLOv5, an object detection model, to identify individual cows in a herd of eight cows captured by utilizing surveillance cameras. We obtained the coordinates of each individual and calculated the distance between them. Next, to extract fighting behaviors related to tank occupancy, we classified and tracked behaviors based on changes in the coordinates of each individual in each video frame. We focused on the time when the distance between individuals became small in the food container.The accuracy of the proposed model showed a high fit rate for all classes. The final estimated rankings were consistent with the expert rankings for seven out of eight animals. This showed that social rankings can be automatically obtained from video images.","PeriodicalId":101653,"journal":{"name":"2021 IEEE 20th International Conference on Cognitive Informatics & Cognitive Computing (ICCI*CC)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121674240","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"From Data to Information Granules: An Environment of Granular Computing","authors":"W. Pedrycz","doi":"10.1109/ICCICC53683.2021.9811327","DOIUrl":"https://doi.org/10.1109/ICCICC53683.2021.9811327","url":null,"abstract":"With enormous amounts of data come opportunities of building models of real-world systems that are instrumental in realizing a plethora of control, prediction, and classification tasks. The interpretability facet of ensuing models becomes highly relevant in light of designing autonomous systems and those constructs supporting human-centric decision-making environments. To transform data to tangible and actionable pieces of knowledge and formulate a problem at hand at a suitable level of abstraction, a convenient way to proceed is to position the problem in the environment of Granular Computing.We advocate that a systematic way of capturing knowledge residing within acquired data and encapsulating such knowledge in the form of interpretable models is supported by a suitable level of abstraction at which the data are to be represented. In turn, we show that an abstraction mechanism is conveniently realized in the form of information granules. Information granules and Granular Computing delivers an operational and flexible setting in which granular models are built and analyzed. A formal characterization of information granules is introduced where they are concisely described as triple (G, I, R) capturing their underlying geometry in the data space (G), information content (I), and representation capabilities of the underlying experimental evidence (R).A suite of design methods transforming data into information granules being articulated in various formal settings (e.g., intervals, fuzzy sets, rough sets) is analyzed and an array of generalizations is discussed (including collaborative ways of building granules in the presence of some auxiliary domain knowledge).In the sequel, it is shown how information granules regarded as functional modules are efficiently used in the construction of a vast array of interpretable models, especially rule-based architectures.","PeriodicalId":101653,"journal":{"name":"2021 IEEE 20th International Conference on Cognitive Informatics & Cognitive Computing (ICCI*CC)","volume":"107 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124040710","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"EEG-based Mental Workload Assessment using a Graph Attention Network","authors":"Hrishikesh Rajasekharan, Shreya Chivilkar, Namrata Bramhankar, Tanushree Sharma, R. Daruwala","doi":"10.1109/ICCICC53683.2021.9811325","DOIUrl":"https://doi.org/10.1109/ICCICC53683.2021.9811325","url":null,"abstract":"Sustained high mental workload (MWL) experienced by operators in high-pressure jobs can compromise their performance, potentially endangering them as well as others. Using electroencephalograms (EEG) to gauge MWL levels is an approach that has been gaining prominence lately. Graph attention networks (GAT) have previously been used to great effect for traffic forecasting, citation networks, etc. In that context, we propose a GAT-based approach for improving the assessment of MWL using EEG signals. We focus on distinguishing EEGs corresponding to a high MWL from the EEGs corresponding to a low MWL and provide a comparative analysis of different features viz. band power, wavelet features, and autoregressive (AR) parameters. The obtained results show that this approach achieves an average accuracy of up to 95.66%, which is superior to that obtained using conventional multilayer perceptron (MLP) and several other recently used methods.","PeriodicalId":101653,"journal":{"name":"2021 IEEE 20th International Conference on Cognitive Informatics & Cognitive Computing (ICCI*CC)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115911024","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}