Retrieving and classifying instances of source code plagiarism

Automatic detection of source code plagiarism is an important research field for both the commercial software industry and within the research community. Existing methods of plagiarism detection primarily involve exhaustive pairwise document comparison, which does not scale well for large software collections. To achieve scalability, we approach the problem from an information retrieval (IR) perspective. We retrieve a ranked list of candidate documents in response to a pseudo-query representation constructed from each source code document in the collection. The challenge in source code document retrieval is that the standard bag-of-words (BoW) representation model for such documents is likely to result in many false positives being retrieved, because of the use of identical programming language specific constructs and keywords. To address this problem, we make use of an abstract syntax tree (AST) representation of the source code documents. While the IR approach is efficient, it is essentially unsupervised in nature. To further improve its effectiveness, we apply a supervised classifier (pre-trained with features extracted from sample plagiarized source code pairs) on the top ranked retrieved documents. We report experiments on the SOCO-2014 dataset comprising 12K Java source files with almost 1M lines of code. Our experiments confirm that the AST based approach produces significantly better retrieval effectiveness than a standard BoW representation, i.e., the AST based approach is able to identify a higher number of plagiarized source code documents at top ranks in response to a query source code document. The supervised classifier, trained on features extracted from sample plagiarized source code pairs, is shown to effectively filter and thus further improve the ranked list of retrieved candidate plagiarized documents.     

Improving cognitive and physical function through 12-weeks of resistance training in older adults: Randomized controlled trial

ABSTRACT

We performed a randomized, controlled trial to analyse the effects of resistance training (RT) on cognitive and physical function among older adults. Fifty participants (mean age 67 years, ~60% woman) were randomly assigned to an RT program or a control group. Participants allocated to RT performed three sets of 10-15RM in nine exercises, three times per week, for 12-weeks. Control group did not perform any exercise. Variables included cognitive (global and executive function) and physical function (gait, mobility and strength) outcomes. At completion of the intervention, RT was shown to have significantly mitigated the drop in selective attention and conflict resolution performance (Stroop test: -494.6; 95%CI: -883.1; ?106.1) and promoted a significant improvement in working memory (digit span forward: -0.6; 95%CI: ?1.0; ?0.1 and forward minus backward: -0.9; 95% CI: ?1.6; ?0.2) and verbal fluency (animal naming: +1.4, 95%CI 0.3, 2.5). No significant between-group differences were observed for other cognitive outcomes. Regarding physical function, at completion of the intervention, the RT group demonstrated improved fast-pace gait performance (?0.3; 95% CI: ?0.6; ?0.0) and 1-RM (+21.4 kg; 95%CI: 16.6; 26.2). No significant between-group differences were observed for other mobility-related outcomes. In conclusion, RT improves cognitive and physical function of older adults.     

The effect of smartphones and playing video games on decision-making in soccer players: A crossover and randomised study

ABSTRACT

The objective of this study was to analyse the effect of the use of social networks in smartphones or playing video games on the passing decision-making performance in professional soccer athletes. Participants were 25 male professional soccer athletes (mean ± SD: age 23.4 ± 2.8 years). The participants performed three randomised conditions divided into three groups: control (CON), smartphone (SMA), and video game (VID). Before and after each experimental condition, the Stroop Task assessed the level of induced mental fatigue. Then, the athletes performed a simulated soccer match. A CANON® camera recorded the matches for further analysis on passing decision-making performance. A group effect was identified (p < .01) with impairment on passing decision-making performance for the SMA (p = .01, ES = 0.5) and VID (p = .01, ES = 0.5) conditions. We concluded that the use of social networks on smartphones and/or playing video games right before official soccer matches might impair the passing decision-making performance in professional soccer athletes.     

Impact of sports participation on mortality rates among Brazilian adults

Objective: To analyze the possible dose-response association between components of sports participation (intensity, volume and previous engagement) and 4-year mortality rates among Brazilian adults.

Methods: 679 males and females (mean age among men = 66.7 ± 9.3 years old and mean age among women = 64.8 ± 8.9 years old) composed the study sample. Sports participation was assessed using Baecke’s questionnaire, which considers intensity, duration and previous engagement. Medical records were used to identify the cause of the death. Cox regression analysis was performed to determine the independent associations of exercise components and mortality.

Results: Participants that reported exercising at moderate-vigorous intensity (Moderate/vigorous: 4.1% versus None/light: 10.3% [p-value = 0.012]; HR = 0.42 [0.1 to 0.94)] and for more than four months (≥4 months: 5.3% versus <4 months: 10.2% [p-value = 0.038]; HR = 0.47 [0.24 to 0.94]) had lower mortality risk. The percentage of survival according to all-cause mortality was significantly higher for participants engaged in sports at moderate-vigorous intensity (p-value = 0.014), as well as for those engaged in sports for periods superior than four months (p-value = 0.036).

Conclusion: We found higher percentage of survival among adults engaged in sports at moderate-vigorous intensity and with at least four months of previous engagement.     

How teacher and students study ‘all that matters’ in high school chemistry

This study focuses on the relationships between meanings of science constructed in classrooms and sociocultural precepts of the wider society. Research presented is part of a larger study of the negotiation of knowledge and roles in high school science classrooms. The data presented document that what goes on in classrooms reflects and recreates societal attitudes towards work, in general, and teachers’ work, in particular. The teacher in this study reduces his demands for academic work in return for students’ cooperation and as a means of coping with negative aspects of teaching. Students, in turn, are not concerned about the specific subject matter of chemistry. What matters to them are the academic credentials that lead to success in the college and employment arenas. As a result, teacher and students construct school chemistry to be what you‐get‐through‐as‐you‐get‐along. The interdependence of the roles and actions of teachers and students suggests changes in policy and practice.     

Biomechanical characteristics of adults walking forward and backward in water at different stride frequencies

The aim of this study was to examine spatiotemporal characteristics and joint angles during forward and backward walking in water at low and high stride frequency. Eight healthy adults (22.1 ± 1.1 years) walked forward and backward underwater at low (50 pulses) and high frequency (80 pulses) at the xiphoid process level with arms crossed at the chest. The main differences observed were that the participants presented a greater speed (0.58 vs. 0.85 m/s) and more asymmetry of the step length (1.24 vs. 1.48) at high frequency whilst the stride and step length (0.84 vs. 0.7 m and 0.43 vs. 0.35 m, respectively) were lower compared to low frequency (P < 0.05). Support phase duration was higher at forward walking than backward walking (61.2 vs. 59.0%). At initial contact, we showed that during forward walking, the ankle and hip presented more flexion than during backward walking (ankle: 84.0 vs. 91.8º and hip: 22.8 vs. 8.0º; P < 0.001). At final stance, the knee and hip were more flexed at low frequency than at high frequency (knee: 150.0 vs. 157.0º and hip: ?12.2 vs. –14.5º; P < 0.001). The knee angle showed more flexion at forward walking (134.0º) than backward walking (173.1º) (P < 0.001). In conclusion, these results show how forward and backward walking in water at different frequencies differ and contribute to a better understanding of this activity in training and rehabilitation.     

Accelerometer data requirements for reliable estimation of habitual physical activity and sedentary time of children during the early years - a worked example following a stepped approach

This study presents a worked example of a stepped process to reliably estimate the habitual physical activity and sedentary time of a sample of young children. A total of 299 children (2.9 ± 0.6 years) were recruited. Outcome variables were daily minutes of total physical activity, sedentary time, moderate to vigorous physical activity and proportional values of each variable. In total, 282 (94%) provided 3 h of accelerometer data on ≥1 day and were included in a 6-step process: Step-1: determine minimum wear-time; Step-2: process 7-day-data; Step-3: determine the inclusion of a weekend day; Step-4: examine day-to-day variability; Step-5: calculate single day intraclass correlation (ICC) (2,1); Step-6: calculate number of days required to reach reliability. Following the process the results were, Step-1: 6 h was estimated as minimum wear-time of a standard day. Step-2: 98 (32%) children had ≥6 h wear on 7 days. Step-3: no differences were found between weekdays and weekend days (P ≥ 0.05). Step-4: no differences were found between day-to-day variability (P ≥ 0.05). Step-5: single day ICC’s (2,1) ranged from 0.48 (total physical activity and sedentary time) to 0.53 (proportion of moderate to vigorous physical activity). Step-6: to reach reliability (ICC = 0.7), 3 days were required for all outcomes. In conclusion following a 7 day wear protocol, ≥6 h on any 3 days was found to have acceptable reliability. The stepped-process offers researchers a method to derive sample-specific wear-time criterion.