Research and Design of the Adaptive Evaluation System Based on the IRT Theory

This paper, on the basis of the author realizing the skill evaluation system based on real environment, discusses several commonly used parameter estimation methods based on item response theory （ IRT ） and analyzes the advantages and disadvantages of each estimation method and their respective application fields. Also, it expounds the research theory and design process of skill adaptive evaluation system based on real environment and the innovation of the system.

Nuclear Fusion Research and Development Need New Relativistic Mass and Energy Corrections Given by the Information Relativity Theory

Hundred years after the conjecture of the British astronomer Eddington that the sun is powered by nuclear fusion of hydrogen, new physics theory may help make energy harvesting by nuclear fusion soon a reality. Researchers as well as investors funding fusion megaprojects are asked to deal with new relativistic corrections for mass and energy proposed by Suleiman in his Information Relativity Theory (IRT). These corrections were calculated in this contribution. It will help to decide whether a venture will be successful and to save big investments when in doubt. The assumed optimal kinetic energy for controlled nuclear fusion must be corrected to a somewhat higher level. At very high kinetic energy in the upper GeV range, it remains not enough baryonic mass to be transformed in energy. The fusion probability faded out to zero already at the golden limit of the recession speed of between target nucleon and projectile nucleon. Cold nuclear fusion, if ever possible, is recommended for protons rather than deuterons at highest experimental possible temperatures around 1000 (K) and needs fine-tuned kinetic nucleon energy. It would be also of interest whether a golden ratio based nuclear fuel confinement chamber could be beneficial. In this connection, also cold nuclear fusion setups should be discussed. Nature is governed by the golden ratio and criticality of physical systems influenced by it, and nuclear physics is not an exception. Computer simulations of the underlying controlled nuclear fusion processes should gain profit from IRT corrected starting information and may tackle anew possible low energy nuclear transmutations considering the wave-like dark components of matter and energy.

Reducing Item Exposure in Computerized Adaptive Testing Systems Using Automatic Item Generation

This paper studies the technics of reducing item exposure by utilizing automatic item generation methods. Known test item calibration method uses item parameter estimation with the statistical data, collected during examinees prior testing. Disadvantage of the mentioned item calibration method is the item exposure; when test items become familiar to the examinees. To reduce the item exposure, automatic item generation method is used, where item models are being constructed based on already calibrated test items without losing already estimated item parameters. A technic of item model extraction method from the already calibrated and therefore exposed test items described, which can be used by the test item development specialists to integrate automatic item generation principles with the existing testing applications.

A Probabilistic Framework for Temporal Cognitive Diagnosis in Online Learning Systems

Cognitive diagnosis is an important issue of intelligent education systems,which aims to estimate students'proficiency on specific knowledge concepts.Most existing studies rely on the assumption of static student states and ig-nore the dynamics of proficiency in the learning process,which makes them unsuitable for online learning scenarios.In this paper,we propose a unified temporal item response theory(UTIRT)framework,incorporating temporality and random-ness of proficiency evolving to get both accurate and interpretable diagnosis results.Specifically,we hypothesize that stu-dents'proficiency varies as a Wiener process and describe a probabilistic graphical model in UTIRT to consider temporali-ty and randomness factors.Furthermore,based on the relationship between student states and exercising answers,we hy-pothesize that the answering result at time k contributes most to inferring a student's proficiency at time k,which also re-flects the temporality aspect and enables us to get analytical maximization(M-step)in the expectation maximization(EM)algorithm when estimating model parameters.Our UTIRT is a framework containing unified training and inferenc-ing methods,and is general to cover several typical traditional models such as Item Response Theory(IRT),multidimen-sional IRT(MIRT),and temporal IRT(TIRT).Extensive experimental results on real-world datasets show the effective-ness of UTIRT and prove its superiority in leveraging temporality theoretically and practically over TIRT.