Search

Natural language processing (NLP) is the best solution to extensive, unstructured, complex, and diverse network big data for counter-terrorism. Through the text analysis, it is the basis and the most critical step to quickly extract the relationship between the relevant entities pairs in terrorism. Relation extraction lays a foundation for constructing a knowledge graph (KG) of terrorism and provides technical support for intelligence analysis and prediction. This paper takes the distant-supervised relation extraction as the starting point, breaks the limitation of artificial data annotation. Combining the Bidirectional Encoder Representation from Transformers (BERT) pre-training model and the sentence-level attention over multiple instances, we proposed the relation extraction model named BERT-att. Experiments show that our model is more efficient and better than the current leading baseline model over each evaluative metrics. Our model applied to the construction of anti-terrorism knowledge map, it used in regional security risk assessment, terrorist event prediction and other scenarios.

The construction of integrated solar storage and charging power stations has become the key issue in the development of new energy. The effects of insufficient power supply, effective charging time, load uncertainty and user evaluation during the operation of charging stations are comprehensively considered in this paper, and a safety evaluation index system based on fuzzy comprehensive evaluation theory is proposed. In order to ensure the accuracy of the comprehensive evaluation results, a two-layer evaluation model is used for fuzzy comprehensive evaluation to avoid too small weight values of evaluation factors. Finally, an example is given to prove the effectiveness of the evaluation model.

Noise exposure and aging can cause subclinical hearing problems that are not identified with conventional audiometry. Potential clinical assessments were studied that could identify subclinical hearing deterioration: (1) Electrocochleography (ECoG); (2) Middle Ear Muscle Reflex (MEMR); (3) Extended High Frequency (EHF) thresholds (10–16 kHz); and (4) word recognition in noise. Sixty-one young, normal-hearing people were recruited. Each completed the four auditory assessments and the Life-Time Exposure to Noise Solvents Questionnaire (LENS-Q). ECoG and MEMR measurements were measured as a function of sound level using clinical equipment. Word recognition in noise was measured with a diotic, closed-set, 12-alternative spondees-in-noise test. It was hypothesized that normal-hearing individuals reporting more noise exposure would have slower growth of the ECoG measures and MEMR magnitudes with increasing levels. Individuals reporting higher noise exposure with the LENS-Q had significantly reduced MEMR magnitudes and smaller slopes, supporting the hypothesis. Gender effects were observed. No significant effect of self-report noise-exposure was seen with ECoG AP growth functions or SP/AP ratios, though modest age effects were observed. EHF thresholds were the only significant predictor of word recognition in noise. The results suggest that word recognition in noise, EHF audiometry, and MEMR could indicate subclinical hearing problems.

Truncated electrodes with different cone angle are widely used in vehicle body assembly. However, the effect of the cone angle on welding process still keeps unclear. In this paper, a multi-physics finite element model, which considered the coupling of electric field, magnetic field, fluid flow and heat transfer, was used to study the effect of the cone angle variation on resistance spot welding. Four kinds of cone angles, including 15°, 30°, 45°, and 60°, were studied systematically. Research results showed that the cone angle variation affects not only weld quality but also electrode life. Specifically, the lower cone angle, the more severe current density singularity at the tip edge, the stronger electromagnetic stirring in the molten nugget, and thus the stronger fluid flow and more uniform temperature field in the weld nugget, which would produce a positive effect on the weld quality. At the same time, the low cone angles would enhance the cooling ability of the electrode for the increased mass and present lower electrode temperature, and thus have a better electrode life, which was also verified with electrode wear experiments. However, the low cone angle electrodes would produce smaller nuggets and consume more coppers. Moreover, for the small cone angle electrode, a larger welding current has to be applied to compensate the heat dissipated by the electrodes due to their excellent cooling capability. As a result, considering the effects of the electrode cone angle on electrode life, weld quality and production cost, the 30° and 45° cone angle would be recommended in vehicle body production.

Implementation of lightweight low-ductility materials such as aluminum alloys, magnesium alloys and composite materials has become urgently needed for automotive manufacturers to improve the competitiveness of their products. However, the hybrid use of these materials poses big challenges to joining processes. Self-piercing riveting (SPR) is currently the most popular technique for joining dissimilar materials and has been widely used in joining all-aluminum and multi-material vehicle bodies. However, in riveting magnesium alloys, cracks always occur for its low ductility. In this paper, a hybrid joining process named friction self-piercing riveting (F-SPR), which combines mechanical joining mechanism of SPR with solid-state joining mechanism of friction stir spot welding (FSSW) by making rivet rotating at high speed in riveting process, was proposed aiming at joining the low ductility materials. 1-mm-thick AA6061-T6 and 2-mm-thick AZ31B were used to validate the effectiveness of the F-SPR process. The results showed that the F-SPR process could significantly improve the rivetability of magnesium alloys, and greatly increase the joint strength, comparing with traditional SPR process.Copyright © 2013 by ASME