Predictive Judgments Research Articles

The Construction of Judgment Model for Cultivating Intelligent Accounting Professionals in Universities Based on Logarithmic Periodic Power Law Model

Abstract The talent training evaluation model not only helps to evaluate the talent itself but also provides feedback on the content of the talent training evaluation. Therefore, this paper establishes an efficient and intelligent talent training evaluation model for accounting professionals based on the logarithmic cycle power law model. The main content of talent training evaluation is set as general knowledge skills, professional thinking, and values. The log-periodic power-law model and the least squares method are combined to reduce the dimensionality of the nonlinear parameters of the judging content and to quantify the judging of intelligent accounting professional talent training in universities, which is convenient for the calculation of linear functions. With the help of log-periodic power-law oscillation to prove that talent training is changing in a cyclical pattern, the feasibility of its prediction is demonstrated. The study shows that the talent cultivation judgment model constructed based on the log-periodic power-law model is very accurate, especially in talent cultivation value judgment prediction. The model achieves zero error in the prediction of some data, and the maximum error between prediction and actual is only 6%. In the judgment of general knowledge and skill cultivation, the maximum error between the prediction and the actual score of the model is no more than 2 points. This shows that the talent development evaluation model based on the log-periodic power law model can make accurate predictions of talent development evaluation.

An Approach Based on Cross-Attention Mechanism and Label-Enhancement Algorithm for Legal Judgment Prediction

Legal Judgment Prediction aims to automatically predict judgment outcomes based on descriptions of legal cases and established law articles, and has received increasing attention. In the preliminary work, several problems still have not been adequately solved. One is how to utilize limited but valuable label information. Existing methods mostly ignore the gap between the description of established articles and cases, but directly integrate them. Second, most studies ignore the mutual constraint among the subtasks, such as logically or semantically, each charge is only related to some specific articles. To address these issues, we first construct a crime similarity graph and then perform a distillation operation to collect discriminate keywords for each charge. Furthermore, we fuse these discriminative keywords instead of established article descriptions into case embedding with a cross-attention mechanism to obtain deep semantic representations of cases incorporating label information. Finally, under a constraint among subtasks, we optimize the one-hot representation of ground-truth labels to guarantee consistent results across the subtasks based on the label-enhancement algorithm. To verify the effectiveness and robustness of our framework, we conduct extensive experiments on two public datasets. The experimental results show that the proposed method outperforms the state-of-art models by 3.89%/7.92% and 1.23%/2.50% in the average MF1-score of the subtasks on CAIL-Small/Big, respectively.

Contrastive Learning for Legal Judgment Prediction

Legal judgment prediction (LJP) is a fundamental task of legal artificial intelligence. It aims to automatically predict the judgment results of legal cases. Three typical subtasks are relevant law article prediction, charge prediction, and term-of-penalty prediction. Due to the wide range of potential applications, LJP has attracted a great deal of interest, prompting the development of numerous approaches. These methods mainly focus on building a more accurate representation of a case’s fact description in order to improve the performance of judgment prediction. They overlook, however, the practical judicial scenario in which human judges often compare similar law articles or possible charges before making a final decision. To this end, we propose a supervised contrastive learning framework for the LJP task. Specifically, we train the model to distinguish (1) various law articles within the same chapter of a Law and (2) similar charges of the same law article or related law articles. By this means, the fine-grained differences between similar articles/charges can be captured, which are important for making a judgment. Besides, we optimize our model by identifying cases with the same article/charge labels, allowing it to more effectively model the relationship between the case’s fact description and its associated labels. By jointly learning the LJP task with the aforementioned contrastive learning tasks, our model achieves better performance than the state-of-the-art models on two real-world datasets.

Mirror, mirror on the wall, who is the fairest dancer of them all? A naturalistic lens model study on the judgment of dance performance

Success as a dancer is closely associated with positive dance judgments by perceivers. Although dancers’ physical appearance (attractiveness, style) might affect dance judgments beyond dance-specific attributes (technique, expression), they have largely been unconsidered in previous studies. To contribute to a comprehensive explanation of real-life dance judgments, we applied the lens model, an approach explicitly developed to explain the emergence of social judgments by multiple attributes. Therefore, video-records of 70 solo performances were (1) rated regarding dancers’ physical appearance, technique, and expression and (2) judged by 33 perceivers. Results of cross-classified mixed-effects models revealed that attributes of all domains were significantly related to dance judgements. Considered simultaneously, however, only dance-specific attributes contributed to the prediction of dance judgments. Additional moderation analyses underscored the importance of perceivers’ expertise in judging dance. We discuss the lens model as suitable framework for a naturalistic approach to the study of aesthetic experiences and sports performances.

FCA-LJP: A Method Based on Formal Concept Analysis for Case Judgment Prediction

FCA-LJP: A Method Based on Formal Concept Analysis for Case Judgment Prediction

Prediction of court decision from Arabic documents using deep learning

AbstractThe increasing amount of electronic legal documents represents a great opportunity for the development of intelligent computational systems for legal texts processing and classification. Most of these systems use classical machine learning and large datasets in English. This paper proposes an approach to automatically predict legal case outcome from written description of the events in Arabic using deep learning. An in‐house corpus from the decisions of the Moroccan Court of Cassation is built and used to train a deep learning model that predicts judgement. As the created corpus is of limited size, a new data augmentation method is proposed to boost the prediction performance. Two settings for text representation are tested, namely FastText and GloVe embeddings, and multiple deep learning models architectures are tested. The proposed approach succeeds in predicting judicial decisions of the Moroccan Court of Cassation with an accuracy of 80.51% on six classes. Even with a small dataset, the proposed data augmentation method was helpful in improving the overall models' performance. Despite the advancement in the area of legal judgement prediction over the years, this work is the first attempt to predict legal outcome using the documents of the Moroccan Cassation court. The corpus created in the context of this work will be made publicly available to the community.

Using Artificial Intelligence Tools in the Judicial Domain and the Evaluation of their Impact on the Prediction of Judgments

In this article, we present a survey of using artificial intelligence tools and measure their impact on the performance of the judicial system. We study its methods, its challenges, and its potential impacts, as well as its famous applications. Our contribution is reviewing the literature on discovery different existing Maching Learning methods that help to obtain pertinent results for the prediction of judgments, and we will then proceed to present a comparative study of some techniques used in the interest of implementing an intelligent system of decision-making for the Moroccan Courts.

Retracted] Construction of Land Ecological Sustainable Evaluation System Based on Dynamic Simulation and TOPSIS Model

Land has always been the resource on which human beings depend for survival and development, which not only provides space for survival but also is the fundamental source of productivity. With the development of commercialization and urbanization, land resources have been destroyed. As a basic living resource, sustainable land use is particularly important. According to the actual situation of regional resource conditions, with the help of TOPSIS model and dynamic simulation, the land ecological evaluation system is established, and the ecological sustainable development is carried out in the whole country, and the research on land greening is deepened. The results show that (1) the economic level of sustainable utilization of land resources has been continuously improved, from 0.15 to 0.38 in the main urban area, from 0.12 to 0.36 in the commercial area, from 0.1 to 0.37 in the suburbs, from 0.11 to 0.38 in the towns, and from 0.25 to 0.5 in the countryside, to provide economic support for ecological protection. (2) According to the use area of relevant land types, forest land and cultivated land occupy the largest with 32830 hm and 13953 hm, respectively, which are the largest usable areas and the most important resources for ecological protection. (3) In the calculation of future ecological construction area by relevant models, the overall coordination control is obtained from the initial high prediction judgment of 0.552 to 0.888. (4) From different development evaluation angles, the economic angle has always provided the greatest support for ecological protection, and its high‐level dependence level value is as high as 0.72, compared with 0.55 and 0.6 from the social angle and environmental angle, so the sustainable ecological protection construction needs a lot of economic support.

Knowledge-Enriched Multi-Cross Attention Network for Legal Judgment Prediction

Knowledge-Enriched Multi-Cross Attention Network for Legal Judgment Prediction

Retracted: Study of Deep Learning-Based Legal Judgment Prediction in Internet of Things Era.

[This retracts the article DOI: 10.1155/2022/8490760.].

A Survey on Legal Judgment Prediction: Datasets, Metrics, Models and Challenges

A Survey on Legal Judgment Prediction: Datasets, Metrics, Models and Challenges

Using attention methods to predict judicial outcomes

The prediction of legal judgments is one of the most recognized fields in Natural Language Processing, Artificial Intelligence, and Law combined. By legal prediction, we mean intelligent systems capable of predicting specific judicial characteristics such as the judicial outcome, the judicial class, and the prediction of a particular case. In this study, we used an artificial intelligence classifier to predict the decisions of Brazilian courts. To this end, we developed a text crawler to extract data from official Brazilian electronic legal systems, consisting of two datasets of cases of second-degree murder and active corruption. We applied various classifiers, such as Support Vector Machines, Neural Networks, and others, to predict judicial outcomes by analyzing text features from the dataset. Our research demonstrated that Regression Trees, Gated Recurring Units, and Hierarchical Attention Networks tended to have higher metrics across our datasets. As the final goal, we searched the weights of one of the algorithms, Hierarchical Attention Networks, to find samples of the words that might be used to acquit or convict defendants based on their relevance to the algorithm.

Does referential expectation guide both linguistic and social constraints on pronoun comprehension?

Current models suggest that pronoun comprehension is guided by expectations about who or what will be mentioned (Arnold, 1998; Kehler & Rohde, 2013; Hartshorne et al., 2015; Brocher et al., 2018), which we call referential predictability. Yet there is disagreement about whether these expectations explain all types of discourse biases, and in particular some scholars suggest that the subject bias is unrelated to referential predictability (Kehler & Rohde, 2013; Fukumura & van Gompel, 2015). Moreover, the role of expectation has not been broadly tested against the numerous constraints known to affect pronoun comprehension, and no study has tested whether expectation is related to social constraints like gaze and pointing. In eight experiments we systematically test how both pronoun comprehension and prediction judgments are influenced by four constraints: (1) the subject bias in joint-action predicates like Ana went hiking with Liz; (2) both the goal and subject biases in transfer predicates (Ana threw the ball to Liz or Ana got the ball from Liz); (3) pointing while gazing, and (4) gazing. We replicate and extend the known effects of these constraints on pronoun comprehension. Critically, we find that most of these constraints also affect prediction judgments, but the subject bias is inconsistent across verb types. Results support models in which referential expectation affects pronoun comprehension.

Comparison of Procalcitonin With Commonly Used Biomarkers and Algorithms for Evaluating Suspected Pediatric Musculoskeletal Infection in the Emergency Department.

It is difficult to distinguish between children with infectious versus noninfectious conditions of the musculoskeletal system during initial evaluation. Clinical predictive algorithms potentially support this effort but not without limitations. Procalcitonin (PCT) has been proposed as a biomarker to help differentiate infection from noninfection. This study evaluates the adoption and utility of PCT during initial infection evaluations and assesses test characteristics of commonly used parameters and algorithms. PCT was introduced for initial laboratory evaluation of the suspected musculoskeletal infection. Prospective enrollment occurred from July 2020 to November 2021 with 3 cohorts established after a retrospective review of final diagnoses at the end of treatment: 1) deep infection, 2) superficial infection, and 3) noninfection. Univariate and multivariate logistic regression analysis of parameters and diagnoses was performed. Test characteristics of individual and aggregated parameters were assessed. Among 258 children evaluated, 188 (72.9%) had PCT drawn during the evaluation. An increase of PCT acquisition from 67.8% to 82.4% occurred over the study timeframe. Eighty-five children were prospectively studied, including those with deep infection (n=21); superficial infection (n=10), and noninfection (n=54). Test characteristics of parameters showed accuracy ranging from 48.2% to 85.9%. PCT >0.1ng/mL independently predicted deep infection in 84.7% of cases, outperforming white blood cell, C-reactive protein (CRP), and absolute neutrophil count. Using study thresholds for CRP, erythrocyte sedimentation rate, PCT, and Temp improved accuracy to 89.4%. PCT is a potentially useful biomarker during the initial assessment of children suspected to have a musculoskeletal infection. Systematic evaluation using a combination of parameters improves the accuracy of assessment and assists predictive judgment under uncertainty. PCT <0.1ng/mL, erythrocyte sedimentation rate <18 mm/hr, CRP <3.3mg/dL, and temperature <37.8°C should reasonably reassure clinicians that deep musculoskeletal infection is less likely, given the high negative predictive value and collective accuracy of these parameters. Level III - Retrospective cohort comparison.

Judicial knowledge-enhanced magnitude-aware reasoning for numerical legal judgment prediction

Judicial knowledge-enhanced magnitude-aware reasoning for numerical legal judgment prediction

Pluralistic language policy and multilingual legal texts in Eritrea

ABSTRACT Eritrea does not have one official language and there is experience in using multiple languages in official government communications, education, media and the legal system. Since Eritrean independence in May 1991, Eritrean laws have been promulgated in a mix of Arabic, English and Tigrinya: some are issued in these three languages; some are issued in Arabic and Tigrinya or Tigrinya and English; and the rest are issued in English or Tigrinya only. Given the continuation of some pre-independence laws, laws written in Amharic and Italian are also applicable. There is also a plan to translate four basic laws (called Codes), circulated in 2015 in Arabic, Tigrinya and English, into the remaining seven Eritrean languages and making them operational. The Eritrean legal system which gives equal weight to laws prepared in, or translated into, multiple languages has created, and will continue to create, a difficult atmosphere of linguistic discrepancy with prejudicial consequences on rights and duties of people and on the expectation for consistency, predictability and uniformity of judgments. This article presents the experiences in Eritrean courts in relation to discrepancies in legal texts issued in different languages and suggests possible solutions.

Artificial Intelligence-Assisted Decision-Making Method for Legal Judgment Based on Deep Neural Network

With the arrival of the third revolution of artificial intelligence, the applications of artificial intelligence in the fields of automatic driving, image recognition, smart home, machine translation, medical services, e-sports, and so on can be seen everywhere, and topics about artificial intelligence are constantly emerging. Since 2017, the discussion on artificial intelligence in the field of law has become more and more active. In this context, the application of artificial intelligence in the field of legal judgment and the hypothetical system based on this technology in court judgment has also become the object of discussion from time to time. In this paper, based on the artificial intelligence decision-making method of the deep neural network, aiming at the three subtasks of legal judgment prediction, namely, crime prediction, law recommendation, and sentence prediction, a multi-task judgment prediction model BERT12multi and a sentence interval prediction model BERT-Text CNN are proposed, which improve the prediction accuracy and adopt the knowledge distillation strategy to compress the model parameters and improve the reasoning speed of the judgment model. Experiments on the CAIL2018 data set show that the performance of the deep neural network model in crime prediction and law recommendation tasks can be significantly improved by adopting the pre training model adaptive training, grouping focus loss, and gradient confrontation training strategies. Using a step-by-step sentence prediction strategy can realize the weight sharing of pre training model and make use of the prediction results of charges and laws in sentence prediction. The recall training-prediction strategy can avoid error accumulation and improve the accuracy of sentence prediction. By integrating the artificial intelligence decision-making method, the case reasoning speed can be greatly improved, the highest compressible model volume can be about 11% of the original one, and the reasoning speed can be increased by about 8 times. At the same time, performance close to that of the deep neural model can be obtained, which is superior to other legal decision prediction models based on word embedding.

Study of Deep Learning-Based Legal Judgment Prediction in Internet of Things Era.

Legal judgment prediction is the most typical application of artificial intelligence technology, especially natural language processing methods, in the judicial field. In a practical environment, the performance of algorithms is often restricted by the computing resource conditions due to the uneven computing performance of the devices. Reducing the computational resource consumption of the model and improving the inference speed can effectively reduce the deployment difficulty of the legal judgment prediction model. To improve the prediction accuracy, enhance the model inference speed, and reduce the model memory consumption, we propose a BERT knowledge distillation-based legal decision prediction model, called KD-BERT. To reduce the resource consumption in the model inference process, we use the BERT pretraining model with lower memory requirements to be the encoder. Then, the knowledge distillation strategy transfers the knowledge to the student model of the shallow transformer structure. Experiment results show that the proposed KD-BERT has the highest F1-score compared with traditional BERT models. Its inference speed is also much faster than the other BERT models.

Russian Bancassurance Perspectives: From Path Dependence to Ecosystem

The aim of the study is to identify the future of Russian bancassurance in the “digital” world based on the formation of a reasonable opinion on its prospects within the framework of path dependence and the ecosystem approach. The relevance of the article is determined by the significance of bancassurance for the development of the insurance sector and the Russian financial market as a whole. The author uses methods such as analysis and synthesis, analogy, induction, descriptive analysis, matrix construction, clustering using graphs, and index method. The study is based on the works of domestic and foreign authors, data from the Bank of Russia on the state of the banking sector of the Russian Federation and key indicators of the insurers; reports of the rating agency “RA Expert” on bancassurance; data of Professional Analysis Systems of Markets and Companies Interfax. Based on the results of quantitative analysis, the author concludes that the key factor determining the development of Russian bancassurance within the framework of path dependence is the role of banks as recipients of cash flows. The author shows that the digital transformation of bancassurance is mainly associated with the digitalization of intermediary relations between the bank, individual client and insurance company. To quantify the level of such transformation, the author proposes indices of digitalization of bancassurance. The author shows that bancassurance can be successfully built into ecosystems due to its peculiarities. However, the conditions for the development of ecosystems in the Russian market can lead to the dissolution of bancassurance in ecosystems and the loss of its significance for the development of the Russian financial market. The presented conclusions are to be discussed taking into account the study of new issues in this problem area, including a study of the mergers and acquisitions market to analyse changes in the ownership structure of banks and insurance companies; a study of corporate bancassurance; broader analysis of quantitative values of bancassurance due to new data; analysis of ecosystem development in the Russian financial market in order to move from predictive judgments to a discussion of real situations.

What Makes Learners Overestimate Their Text Comprehension? The Impact of Learner Characteristics on Judgment Bias

It is frequently assumed that learner characteristics (e.g., reading skill, self-perceptions, optimism) account for overestimations of text comprehension, which threaten learning success. However, previous findings are heterogenous. To circumvent a key problem of previous research, we considered cognitive, metacognitive, motivational, and personality characteristics of learners (N = 255) simultaneously with regard to their impact on the judgment biases in prediction and postdiction judgments about factual and inference questions. The main results for the factual questions showed that men, lower reading skill, working memory capacity, and topic knowledge, yet higher self-perceptions of cognitive and metacognitive capacities yielded stronger overestimations for prediction judgments. For inference questions, a lower reading skill, higher self-perceptions of metacognitive capacities, and a higher self-efficacy were related to stronger overestimations for prediction and postdiction judgments. A higher openness was a risk factor for stronger overestimations when making predictions for the inference questions. The findings demonstrate that learner characteristics are a relevant source of judgment bias, which should be incorporated explicitly in theories of judgment accuracy. At the same time, fewer learner characteristics were actually relevant than previous research suggests. Moreover, which learner characteristics impact judgment bias also depends on task requirements, such as factual versus inference questions.