Context-dependent Models Research Articles

Fake news detection: Taxonomy and comparative study

The proliferation of social networks has presented a significant challenge in combating the pervasive issue of fake news within modern societies. Due to the large amount of information and news produced daily in text, audio, and video, the validation and verification of this information have become crucial tasks. Leveraging advancements in artificial intelligence, distinguishing between fake news and factual information through automatic fake news detection systems has become more feasible. Automatic fake news detection has been explored from diverse perspectives, employing various feature extraction and classification models. Nonetheless, empirical evaluations, categorization, and comparisons of existing techniques for handling this problem remain limited. In this paper, we revisit the definitions and perspectives of fake news and propose an updated taxonomy for the field based on multiple criteria: (1) Type of features used in fake news detection; (2) Fake news detection perspectives; (3) Feature representation methods; and (4) Classification approaches. Moreover, we conduct an extensive empirical study to evaluate several feature representation techniques and classification approaches based on accuracy and computational cost. Our experimental results demonstrate that the optimal feature extraction techniques vary depending on the characteristics of the dataset. Notably, context-dependent models based on transformer models consistently exhibit superior performance. Additionally, employing transformer models as feature extraction methods, rather than solely fine-tuning the network for the downstream task, improves overall performance. Through extensive error analysis, we identify that a combination of feature representation methods and classification algorithms, including classical ones, offer complementary aspects and should be considered for achieving better generalization performance while maintaining a relatively low computational cost. For further details, including source codes, figures, and datasets, please refer to our project’s GitHub repository: [https://github.com/FFarhangian/Fake-news-detection-Comparative-Study].

Context-aware deep learning with dynamically assembled weight matrices

Deep neural networks are static by nature, meaning they use a single set of parameters to process each data sample. However, for more complex and larger systems, for which samples can be obtained under a large variety of circumstances, a need for more dynamic networks that adapt to these variations seems apparent. To this end, the concept of context information and its integration in deep learning is considered. Contrary to current practices, that treat all modalities as equally informative to the decision process, contextual and feature information is considered to be vastly different in nature. A set of definitions and subsequent arguments are given in order to provide the necessary clarification regarding the interpretation of context with the purpose of using all information in a maximally efficient way. From this interpretation, a problem statement of context aware deep learning is constructed and consequently linked to its multiple model and transfer learning solutions. These solutions, however, are in-efficient since samples are spread across models. Based on the existence of this multiple model solution, a new approach, which integrates contextual information directly into a single context-dependent model, is proposed. This single model uses weight matrices that are dynamically assembled based on the contextual information to process each data sample individually. This allows for the single model to consume and learn from all samples. The corresponding training routine is constructed and evaluated on multiple benchmark problems. We start with an artificially generated problem on which the methods’ ability to model multiple linear classification problems concurrently is confirmed. Next, both a time series forecasting and image classification dataset are used for evaluation. Evaluations of our proposed method are done and compared to standard context aware implementations based on concatenation and gating. These standard methods implement context information by adding additional parameters in order to try modeling all interactions between the context information and the samples. However, our proposed approach integrates the contextual information directly into the network weights, allowing parameter efficient modeling of dynamic contextual behavior. In both cases the proposed solution outperforms its standard counterparts with significant margins in both evaluation metrics and parameter efficiency. Specifically, a mean absolute error improvement of eleven standard deviations and an eight percent increase in classification accuracy, for the forecasting and image classification problems respectively, is observed, showing the potential of our approach.

On Fuzzy and Case-Based Dynamic Software Development Process Modeling and Simulation Approach

The software development process (SDP) is complex because it uses specific knowledge, skills, tools, and techniques to deliver value to people, and the user’s needs are always growing and becoming more complex. Practitioners and academics have proposed different approaches for improving the successful implementation of the SDP. This remains a non-trivial task, and the number of successful software development projects increases only slightly each year. This paper proposes a new approach for adapting the classical SDP to a dynamic, knowledge-intensive, and context-dependent model by applying the case-handling and fuzzy inference approaches for SDP modeling and simulation. The advantages and novelties of this approach consist in: SDP dynamicity modeling using the case-handling approach; the inclusion of human factor uncertainty into the execution of SDP activities through fuzzification; and agent-based simulation, which allows for the inclusion of human actors in SDP simulations. Modeling and simulation of Agile SDPs were carried out using this prototype. Experiments show the correspondence of the proposed approach to the needs of dynamically changing and human-factor-dependent SDPs, as well as the possibilities for modeling and simulating the SDP, predicting SDP execution, and observing possible risks.

A deep learning model for drug screening and evaluation in bladder cancer organoids.

Three-dimensional cell tissue culture, which produces biological structures termed organoids, has rapidly promoted the progress of biological research, including basic research, drug discovery, and regenerative medicine. However, due to the lack of algorithms and software, analysis of organoid growth is labor intensive and time-consuming. Currently it requires individual measurements using software such as ImageJ, leading to low screening efficiency when used for a high throughput screen. To solve this problem, we developed a bladder cancer organoid culture system, generated microscopic images, and developed a novel automatic image segmentation model, AU2Net (Attention and Cross U2Net). Using a dataset of two hundred images from growing organoids (day1 to day 7) and organoids with or without drug treatment, our model applies deep learning technology for image segmentation. To further improve the accuracy of model prediction, a variety of methods are integrated to improve the model's specificity, including adding Grouping Cross Merge (GCM) modules at the model's jump joints to strengthen the model's feature information. After feature information acquisition, a residual attentional gate (RAG) is added to suppress unnecessary feature propagation and improve the precision of organoids segmentation by establishing rich context-dependent models for local features. Experimental results show that each optimization scheme can significantly improve model performance. The sensitivity, specificity, and F1-Score of the ACU2Net model reached 94.81%, 88.50%, and 91.54% respectively, which exceed those of U-Net, Attention U-Net, and other available network models. Together, this novel ACU2Net model can provide more accurate segmentation results from organoid images and can improve the efficiency of drug screening evaluation using organoids.

Enabling Inference for Context-Dependent Models of Mutation by Bounding the Propagation of Dependency.

Although the rates at which positions in the genome mutate are known to depend not only on the nucleotide to be mutated, but also on neighboring nucleotides, it remains challenging to do phylogenetic inference using models of context-dependent mutation. In these models, the effects of one mutation may in principle propagate to faraway locations, making it difficult to compute exact likelihoods. This article shows how to use bounds on the propagation of dependency to compute likelihoods of mutation of a given segment of genome by marginalizing over sufficiently long flanking sequence. This can be used for maximum likelihood or Bayesian inference. Protocols examining residuals and iterative model refinement are also discussed. Tools for efficiently working with these models are provided in an R package, which could be used in other applications. The method is used to examine context dependence of mutations since the common ancestor of humans and chimpanzee.

Context dependency of nucleotide probabilities and variants in human DNA

BackgroundGenomic DNA has been shaped by mutational processes through evolution. The cellular machinery for error correction and repair has left its marks in the nucleotide composition along with structural and functional constraints. Therefore, the probability of observing a base in a certain position in the human genome is highly context-dependent.ResultsHere we develop context-dependent nucleotide models. We first investigate models of nucleotides conditioned on sequence context. We develop a bidirectional Markov model that use an average of the probability from a Markov model applied to both strands of the sequence and thus depends on up to 14 bases to each side of the nucleotide. We show how the genome predictability varies across different types of genomic regions. Surprisingly, this model can predict a base from its context with an average of more than 50% accuracy. For somatic variants we show a tendency towards higher probability for the variant base than for the reference base. Inspired by DNA substitution models, we develop a model of mutability that estimates a mutation matrix (called the alpha matrix) on top of the nucleotide distribution. The alpha matrix can be estimated from a much smaller context than the nucleotide model, but the final model will still depend on the full context of the nucleotide model. With the bidirectional Markov model of order 14 and an alpha matrix dependent on just one base to each side, we obtain a model that compares well with a model of mutability that estimates mutation probabilities directly conditioned on three nucleotides to each side. For somatic variants in particular, our model fits better than the simpler model. Interestingly, the model is not very sensitive to the size of the context for the alpha matrix.ConclusionsOur study found strong context dependencies of nucleotides in the human genome. The best model uses a context of 14 nucleotides to each side. Based on these models, a substitution model was constructed that separates into the context model and a matrix dependent on a small context. The model fit somatic variants particularly well.

Обобщенная контекстно-зависимая теоретико-графовая модель фольклорных и литературных текстов

One of the problems of automatic text processing is their attribution. This term is understood as the establishment of the attributes of a text work (determination of authorship, time of creation, place of recording, etc.). The article presents a generalized context-dependent graph-theoretic model designed for the analysis of folklore and literary texts. The minimal structural unit of the model (primitive) is a word. Sets of words are combined into vertices, and the same word can be related to different vertices. Edges and graph substructures reflect the lexical, syntactic and semantic links of the text. The characteristics of the model are its fuzziness, hierarchy and temporality. As examples, a hierarchical graph-theoretical model of components (on the example of literary works by A. S. Pushkin), a temporal graph-theoretic model of a fairy tale plot (on the example of Russian fairy tales by A. M. Afanasyev) and a fuzzy graph-theoretic model of «strong» connections of grammatical classes (on the example of anonymous articles from the pre-revolutionary magazines «Time», «Epoch» and the weekly «Citizen», edited by F. M. Dostoevsky). The model is built in such a way that it can be further explored using artificial intelligence methods (for example, decision trees or neural networks). For this purpose, a format for storing such data was implemented in the information system «Folklore», as well as procedures for entering, editing and analyzing texts and their graph-theoretic models.

Glandular Cell Segmentation Method Combined with Dense Connective Blocks and Self-Attention Mechanism

<p indent=0mm>Currently used cell segmentation methods are easily to cause the problem of missegmentation and impreciseness for glandular cell segmentation. A glandular cell segmentation model based on U-Net network is proposed which combines dense connective blocks and self-attention mechanism. Firstly, the convolution layers in the U-Net structure are combined to form the dense connective blocks, so that the information can be extracted from the image at different scales. Then the self-attention mechanism is introduced at the decoder to establish a rich context-dependent model for local features to suppress unnecessary feature propagation and improve the accuracy of glandular cell segmentation. The experimental results on the 2015 MICCAI Gland Segmentation Challenge dataset show that the proposed model, with a small number of extra parameters, can achieve improved performance in terms of <italic>F</italic>₁-score, Mean Dice coefficient, and Hausdorff distance compared with other U-Net based methods.

The default mode network: where the idiosyncratic self meets the shared social world.

The default mode network (DMN) is classically considered an 'intrinsic' system, specializing in internally oriented cognitive processes such as daydreaming, reminiscing and future planning. In this Perspective, we suggest that the DMN is an active and dynamic 'sense-making' network that integrates incoming extrinsic information with prior intrinsic information to form rich, context-dependent models of situations as they unfold over time. We review studies that relied on naturalistic stimuli, such as stories and movies, to demonstrate how an individual's DMN neural responses are influenced both by external information accumulated as events unfold over time and by the individual's idiosyncratic past memories and knowledge. The integration of extrinsic and intrinsic information over long timescales provides a space for negotiating a shared neural code, which is necessary for establishing shared meaning, shared communication tools, shared narratives and, above all, shared communities and social networks.

Context-dependent model for spam detection on social networks

Social media platforms are getting an important communication medium in our daily life, and their increasing popularity makes them an ideal platform for spammers to spread spam messages, known as spam problems. Moreover, messages on social media are vague and messy, so a good representation of the text may be the first step to address spam problem. While traditional weighting methods suffer from both high dimensionality and high sparsity problems, traditional word embedding methods suffer from context independence and out of vocabulary problems. To overcome these problems, in this study, we propose a novel architecture based on a context-dependent representation of text using the BERT model. The model was tested using the Twitter dataset, and experimental results show that the proposed method outperforms traditional weighting methods, traditional word embedding based methods as well as the existing state of the art methods used to detect spam on the twitter platform.

Social Image Concern and Reference Point Formation

Abstract This paper develops an axiomatic context-dependent model of social image concerns. Allowing for context-dependence based on choice sets, it examines how context-dependence impacts social image concerns, in particular how a decision maker exhibits various social emotions stemming from their intrinsic reference point, which may not be captured by conforming to social norms. To elicit the intrinsic reference point, this paper provides weaker versions of Strategic Rationality and Independence, in addition to the basic axioms, to characterize the model of the Reference-Dependent Image-Conscious utilitarian. This paper also examines how social emotions stemming from the intrinsic reference point are related to preference reversals as violations of the Weak Axiom of Revealed Preference (WARP). Finally, this paper considers the relationship between social image concerns and intrinsic reciprocity. The findings demonstrate that social image plays a large role and the model developed illustrates a condition in which cooperation is sustained in prisoner’s dilemma games.

Syllable based Hindi speech recognition

In this paper, one of the acoustic units of speech, the syllable, is used for the development of a continuous Hindi speech recognition system. The syllable is a larger acoustic unit that overcomes the contextual effects and requires fewer training samples in comparison to triphone based and word-based models. Other acoustic units such as phoneme-based suffer from contextual influences, and context-dependent triphones suffer due to the non-availability of triphone patterns with a large memory storage for numerous models. Earlier research works related to Hindi speech recognition were performed using the word, phoneme, and context-dependent models. The authors proposed a syllable based Hindi speech recognition system in this study due to different advantages of syllable units such as longer acoustic units, fast decoding, reducing contextual effects, and reduction of irregularities due to phonemes. The continuous Hindi speech recognition system was developed utilizing syllable based acoustic units. Hindi is widely spoken in India and other parts of the world also. The experiments are performed on Continuous Hindi speech by using a widely known Hidden Markov Model (HMM) with perceptual linear predictive coefficients(PLPs). The research outcomes reveal that by using syllables, the performance of the system was increased by 27% than phoneme and 20% than triphones. Research findings indicate that by selecting an appropriate acoustic unit for Hindi, the performance of the speech recognition system may be improved. Further, the study also provides useful insights to develop a syllable based pronunciation dictionary that may be used in speech recognition, speaker identification, and text to speech conversion systems.

EvoLSTM: context-dependent models of sequence evolution using a sequence-to-sequence LSTM.

MotivationAccurate probabilistic models of sequence evolution are essential for a wide variety of bioinformatics tasks, including sequence alignment and phylogenetic inference. The ability to realistically simulate sequence evolution is also at the core of many benchmarking strategies. Yet, mutational processes have complex context dependencies that remain poorly modeled and understood.ResultsWe introduce EvoLSTM, a recurrent neural network-based evolution simulator that captures mutational context dependencies. EvoLSTM uses a sequence-to-sequence long short-term memory model trained to predict mutation probabilities at each position of a given sequence, taking into consideration the 14 flanking nucleotides. EvoLSTM can realistically simulate mammalian and plant DNA sequence evolution and reveals unexpectedly strong long-range context dependencies in mutation probabilities. EvoLSTM brings modern machine-learning approaches to bear on sequence evolution. It will serve as a useful tool to study and simulate complex mutational processes.Availability and implementationCode and dataset are available at https://github.com/DongjoonLim/EvoLSTM.Supplementary information Supplementary data are available at Bioinformatics online.

DNA Hypermethylation and Unstable Repeat Diseases: A Paradigm of Transcriptional Silencing to Decipher the Basis of Pathogenic Mechanisms.

Unstable repeat disorders comprise a variable group of incurable human neurological and neuromuscular diseases caused by an increase in the copy number of tandem repeats located in various regions of their resident genes. It has become clear that dense DNA methylation in hyperexpanded non-coding repeats induces transcriptional silencing and, subsequently, insufficient protein synthesis. However, the ramifications of this paradigm reveal a far more profound role in disease pathogenesis. This review will summarize the significant progress made in a subset of non-coding repeat diseases demonstrating the role of dense landscapes of 5-methylcytosine (5mC) as a common disease modifier. However, the emerging findings suggest context-dependent models of 5mC-mediated silencing with distinct effects of excessive DNA methylation. An in-depth understanding of the molecular mechanisms underlying this peculiar group of human diseases constitutes a prerequisite that could help to discover novel pathogenic repeat loci, as well as to determine potential therapeutic targets. In this regard, we report on a brief description of advanced strategies in DNA methylation profiling for the identification of unstable Guanine-Cytosine (GC)-rich regions and on promising examples of molecular targeted therapies for Fragile X disease (FXS) and Friedrich ataxia (FRDA) that could pave the way for the application of this technique in other hypermethylated expansion disorders.

Intrapersonal heterogeneity in car-sharing decision-making processes by activity-travel contexts: A context-dependent latent class random utility – random regret model

This paper investigates stated preferences for car-sharing in the context of travel mode choice under conditions of uncertain travel times. In particular, the study focuses on heterogeneity in the decision-making process due to different activity-travel contexts such as time pressure, activity duration, and uncertain travel times. It is assumed that the contextual factors influence the decision rules that potential users adopt regarding the use of a shared car for a specific trip purpose. A context-dependent latent class model allowing for the random panel effect is estimated, in which the latent classes respectively represent utility maximization and regret minimization decision rules. The model employs the contextual factors as covariates of class membership probability. A stated choice experiment, in which car-sharing, own car and public transport serve as the choice alternatives, is used to estimate the model. Results indicate that the activity-travel contexts play an important role in accounting for heterogeneity in decision rules.

Context-Dependent Sequence-to-Sequence Turkish Spelling Correction

In this article, we make use of sequence-to-sequence (seq2seq) models for spelling correction in the agglutinative Turkish language. In the baseline system, misspelled and target words are split into their letters and the letter sequences are fed into the seq2seq model. We prefer letters as the unit of the model due to the agglutinative nature of Turkish, which results in an impractical dictionary size when words are used as a dictionary unit. In order to improve the baseline performance, we incorporate right and left context of the misspelled words. All context words are represented with their first three consonants in the context-dependent model. We train the seq2seq models using a large text corpus collected automatically from the Internet. The corpus contains approximately 4 million sentences. We randomly introduce substitution, deletion, and insertion spelling errors to the words in the corpus. We test the performance of the proposed context-dependent seq2seq model using synthetic and realistic test sets. The synthetic test set is constructed similar to the training set. The realistic test set contains human-made misspellings from Twitter messages. In the experiments, we observed that the proposed context-dependent model performs significantly better than the baseline system. Its correction accuracy reaches 94% on the synthetic dataset. Additionally, the proposed method provides 2.1% absolute improvement over a state-of-the-art Turkish spelling correction system on the Twitter test set.

A Systematic Literature Study to Unravel Transparency Enabled by Open Government Data: The Window Theory

The opening of data has been credited for improving transparency and for providing a window on government functioning. Although this relationship is intuitively apparent, it is in fact complex and the mere opening of data might not actually yield transparency. In this paper, a comprehensive model of determinants that enable or impede transparency enabled by open government data and the expected effects have been derived by surveying public administration and information systems literature. Public administration literature tends to be focused on factors such as participation and trust, whereas information systems literature focuses on factors such as user interface, user experience, and data quality. Digital government literature attempts to bridge these elements. The Window Theory is introduced, in order to unify existing models by integrating a broad range of factors within a single model. The Window Theory can be used to develop context-dependent models that are both comprehensive and parsimonious.

Continuous Speech Recognition System for Kannada Language with Triphone Modelling using HTK

Abstract: Kannada is the regional language of India spoken in Karnataka. This paper presents development of continuous kannada speech recognition system using monophone modelling and triphone modelling using HTK. Mel Frequency Cepstral Coefficient (MFCC) is used as feature extractor, exploits cepstral and perceptual frequency scale leads good recognition accuracy. Hidden Markov Model is used as classifier. In this paper Gaussian mixture splitting is done that captures the variations of the phones. The paper presents performance of continuous Kannada Automatic Speech Recognition (ASR) system with respect to 2, 4,8,16 and 32 Gaussian mixtures with monophone and context dependent tri-phone modelling. The experimental result shows that good recognition accuracy is achieved for context dependent tri-phone modelling than monophone modelling as the number Gaussian mixture is increased.

Standard Yorùbá context dependent tone identification using Multi-Class Support Vector Machine (MSVM)

Most state-of-the-art large vocabulary continuous speech recognition systems employ context dependent (CD) phone units, however, the CD phone units are not efficient in capturing long-term spectral dependencies of tone in most tone languages. The Standard Yorùbá (SY) is a language composed of syllable with tones and requires different method for the acoustic modeling. In this paper, a context dependent tone acoustic model was developed. Tone unit is assumed as syllables, amplitude magnified difference function (AMDF) was used to derive the utterance wide F contour, followed by automatic syllabification and tri-syllable forced alignment with speech phonetization alignment and syllabification SPPAS tool. For classification of the context dependent (CD) tone, slope and intercept of F values were extracted from each segmented unit. Supervised clustering scheme was utilized to partition CD tri-tone based on category and normalized based on some statistics to derive the acoustic feature vectors. Multi-class support vector machine (MSVM) was used for tri-tone training. From the experimental results, it was observed that the word recognition accuracy obtained from the MSVM tri-tone system based on dynamic programming tone embedded features was comparable with phone features. A best parameter tuning was obtained for 10-fold cross validation and overall accuracy was 97.5678%. In term of word error rate (WER), the MSVM CD tri-tone system outperforms the hidden Markov model tri-phone system with WER of 44.47%.Keywords: Syllabification, Standard Yorùbá, Context Dependent Tone, Tri-tone Recognition

Detection and correction of subtle context-dependent robot model inaccuracies using parametric regions

Autonomous robots frequently rely on models of their sensing and actions for intelligent decision making. Unfortunately, in complex environments, robots are bound to encounter situations in which their models do not accurately represent the world. Furthermore, these context-dependent model inaccuracies may be subtle, such that multiple observations may be necessary to distinguish them from noise. This paper formalizes the problem of detection and correction of such subtle contextual model inaccuracies in autonomous robots, and presents an algorithm to address this problem. The solution relies on reasoning about these contextual inaccuracies as parametric regions of inaccurate modeling (RIMs) in the robot’s planning space. Empirical results from various real robot domains demonstrate that, by explicitly searching for RIMs, robots are capable of efficiently detecting subtle contextual model inaccuracies, which in turn can lead to task performance improvement.