Using Deep Learning to Improve Automated Program Static Assessment

Evaluation is an essential part of the teaching process,especially in the programming course.Both students and teachers can benefit significantly from automatic program evaluation.It shortens the time required for assessment so that students can get immediate feedback.At the same time,it can also significantly reduce the workload of teachers.Currently,the automated program assessment system mainly uses a combination of static and dynamic analysis methods.The system is faced with two crucial problems of the unfinished code evaluation and the template code construction.This paper proposes a method of combining deep learning with static analysis.The syntax tree repair is used to solve the problem that the code with compiling errors cannot generate the correct syntax tree.Moreover,the target code is converted to a subset of solution space through the syntax tree standardization,which reduces the number of template code needed.Based on deep learning,the embedded token vector keeps the code’s context all the time,which ensures that the lexical-semantic remains unchanged as much as possible after the syntax tree changes.Finally,the standardized tree is represented as a vector by the recursive neural network.Cosine similarity between target and template code vectors is used as an evaluation score.The experiment shows that the similarity scores obtained by this method are consistent with the expert scores.This method can provide support for future research,such as difficult feedback and has great significance.

Assessing the Effectiveness of a Cervical Cancer Screening Program in a Hospital-based Study

This study compared HPV testing and liquid-based cytology （LCT） as performance indicators for cervical cancer screening in a hospital-based study. A total of 61,193 outpatients were screened initially by LCT. Samples with screening results showing atypical squamous cells of undetermined significance （ASC-US） or worse were referred for colposcopy, and some samples were tested for high-risk HPV types with the Hybrid Capture II system （HC II）. Data on LCT （n=61,193） and HC II （n=1056） results were analysed. Overall test positivity for LCT was 2.53% using an ASC-US threshold, 3.11% using a low-grade squamous intraepithelial lesion （LSIL） threshold, and 0.67% using a high-grade squamous intraepithelial lesion （HSIL） threshold. A total of 1839 women （84% of the 3893 patients with abnormal cytology） underwent colposcopy-directed biopsy. HPV was positive in 80.3% of women with cervical intraepithelial neoplasia 1 （CIN1）, 88.3% of those with CIN2, 79.2% of women with CIN3 and 50% （2 of 4） of women with invasive cancer. There was a significant increase in the detection of CIN2 or worse with adjunct HPV testing of women with ASC-US and LSIL However, there were detection of CIN2＋ cases no differences in the with adjunct HPV testing of women with HSlI.. The results indicate that HPV testing for HSlL triage should not be recommended in cervical cancer screening.

An Exploration of Reasons for Shanghai’s Success in the OECD Program for International Student Assessment(PISA)2009

The outstanding performance of Shanghai students in the 4th Programme for International Student Assessment(PISA 2009)gained widespread attention at home and abroad.In this paper,the authors attribute this outstanding performance to three traditional factors and six modern factors.The traditional factors are high parental expectations,belief in the power of effort,and the mechanisms of personnel selection.The modern factors are the openness of Chinese education,curriculum and teaching reform,teacher training,improvement of comparatively poor schools,the financial resources allocation mechanism in compulsory education,and the reform of high school enrollment.

Performance Analysis and Future Prospect of China’s Anti-bullying Policy and Practice:Evidence from PISA 2015 and PISA 2018

This paper analyzes the performance of China’s anti-bullying efforts with the data on school bullying in the Program for International Student Assessment(PISA)2015 and PISA 2018.It is found that China has fully demonstrated its high performance in anti-bullying actions with all three key performance indicators dropping dramatically,namely the prevalence of bullying,the index of exposure to bullying,and the percentage of frequently bullied students.Further,from the perspectives of personal characteristics and school features,two rounds of PISA data are adopted to make the longitudinal diachronic comparison and the cross-national horizontal comparison,followed by the analysis of the occurrence and development patterns of school bullying.Five suggestions are put forward for the deployment of China’s anti-bullying action plan.

Experience and Challenge of Basic Education Development in China:Reflections on PISA 2018 Results

The outstanding performance of students in four Chinese provinces and municipalities in the Program for International Student Assessment(PISA)2018 is inextricably linked to local resource support,excellent education traditions,and ongoing curriculum and teaching reform.However,deepening education reform requires attention to existing challenges,including large inter-school gaps,the shortage of teachers in urban and rural areas,as well as students’heavy schoolwork burden and low level of well-being.Therefore,several policy suggestions are proposed accordingly:gaining insight into and developing higher quality and fairer education,pursuing effective actions of education reform,and exploring innovative pathways to address educational problems.

Assessing the Quality and Equality of Hong Kong Basic Education Results from PISA 2000+to PISA 2006

This paper seeks to examine the quality and equality of basic education of Hong Kong based on the first three cycles of the Program for International Student Assessment(PISA).Results from these three assessments suggested that the Hong Kong students have an outstanding performance in mathematics,science and reading.In particular,the performance of reading improves substantially in PISA 2006.As far as equality in education is concerned,the achievement gap of students from different socio-economic backgrounds in Hong Kong is relatively small compared with other countries.However,the academic performance variation between-schools suggest that,there is still academic segregation among secondary schools in Hong Kong although it has been reduced slightly in the PISA 2006.

Heterogeneity Analysis on the Influence of the Supply of Teachers on the Academic Performance of 15-Year-Old Students:Based on PISA 2018 Samples from Four Provinces and Municipalities in China

The quantitative and qualitative supply of teachers has a significant influence on student academic performance.Based on the samples collected from four provinces and municipalities in China for the Program for International Student Assessment(PISA)2018,this study analyzes the effects of the supply of teachers on the academic performance of 15-year-old students through a hierarchical linear model(HLM)as well as a quantile regression model.The study has found that the shortage of teachers and the proportion of teachers with a master’s degree significantly affect student academic performance,and the supply of teachers with higher degrees has a decreasing marginal benefit on student academic performance;its best marginal benefit on student academic performance is achieved when the proportion of teachers with a master’s degree reaches about 75%.For students in different academic performance subgroups,the marginal benefit of an increase in the proportion of teachers with a master’s degree in schools is higher for“top students”than for“students with learning difficulties.”In terms of the difference between urban and rural areas,the shortage of teachers in rural schools is more prominent than that in urban schools,and teachers with higher degrees tend to cluster in core cities.The marginal benefit of an increase in the proportion of teachers with a master’s degree is higher for improving student academic performance in rural areas.Therefore,it is recommended to design a policy of differentiated compensatory earnings for rural teachers,and efforts should be enhanced to promote programs for supporting high-quality teachers.In addition,the allocation of teachers should follow the principle of“differentiated equity,”and more differentiated teaching strategies should be adopted for students with different academic abilities.

“Screening”or“Meritocracy”:Does the Choice of a Private School Help Students Achieve Higher Academic Performance?-Based on the PISA 2018 Analysis of Four Provinces and Municipalities in China

The introduced policy of“synchronized enrollment of public and private schools”has once again triggered the debate on the effectiveness of public and private schools.Based on the data of the Program for International Student Assessment(PISA)2018 from four Chinese provinces and municipalities,this paper explores whether private schools gain a relative advantage in student academic performance through student“screening”or academic“meritocracy,”through a hierarchical linear model(HLM)and an empirical test of the propensity score matching(PSM).It has been found that the academic performance of students in private schools is significantly better than that in public schools.But with background,metacognitive ability,and learning hours of students in private schools controlled for,such academic performance is not significantly superior,suggesting that private schools rely heavily on student“screening”to achieve a relative advantage in student academic performance.This finding has also verified the scientific nature of the above policy.

How the ICT Development Level Influences Students’Digital Reading Literacy:A Multi-level Model Comparison Based on PISA 2018 Data

The purpose of this study is to systematically showcase and evaluate how students’,schools’,and countries’Information and Communication Technology(ICT)development levels influence students’digital reading literacy by using the data of the Program for International Student Assessment(PISA)2018 and the multi-level model approach.It is found that the ICT development at each level has a significant positive effect on students’digital reading literacy,and there is a significant interactive moderating effect between different levels.The gap in ICT development levels between the countries participating in the Belt and Road Initiative(BRI)and developed economies has a“Matthew effect”in widening the digital reading literacy gap in the short term,but in the long run,the faster growth of the countries participating in BRI in ICT development levels will narrow the gap and alleviate the effect.China should make use of its advantages in ICT development to engage in global education governance,facilitate the co-development in both the ICT development level and the digital reading literacy for the countries participating in BRI,and contribute to the development of global education by opening up a new way of development.