Search

XPath query is the key part of XML data processing, and its performance is usually critical for XML applications. In the process of XPath query, there is inherent seriality between query steps, which makes it difficult to parallelize the query effectively as a whole. On the other hand, although XPath query has the characteristics of data stream processing and is suitable for pipeline processing, the data flow of each query step usually varies a lot, which results in limited performance under multithreading conditions. In this paper, we propose a pipelined XPath query method (PXQ) based on cost optimization. This method uses pipelined query primitives to process query steps based on relation index. During pipeline construction, a cost estimation model based on XML statistics is proposed to estimate the cost of the query primitive and provide guidance for the creation of a pipeline phase through the partition of query primitive sequence. The pipeline construction technique makes full use of available worker threads and optimizes the load balance between pipeline stages. The experimental results show that our method can adapt to the multithreaded environment and stream processing scenarios of XPath query, and its performance is better than the existing typical query methods based on data parallelism. [ABSTRACT FROM AUTHOR]

The stock turnover values are sensitive to external factors, and remain great challenges in its prediction. The consideration is that search engine data can reflect market environment, policies and attentions on stocks. Therefore, a dual sides autoregression (DSAR) method is proposed to benefit from both observed turnover values and exogenous data. The proposed DSAR consists of linear representation stage and combination stage. In linear representation stage, the short-term patterns of turnover values and query data are represented, respectively. In combination stage, the outputs from previous stages are combined. Intensive experiments on two groups of data collections show the effectiveness of our proposed method. [ABSTRACT FROM AUTHOR]

The LiDAR point cloud filtering algorithm based on marginal points, is performed well in complex cityscapes. However, it is very time consuming for detecting marginal points. To improve the efficiency for LiDAR point cloud filtering, a parallel marginal point detecting algorithm is proposed based on GPGPU. The algorithm is followed by four steps: parallel indexing K nearest neighbor points, parallel plane fitting, parallel calculating projection points and parallel calculating points' angle. For the step of parallel indexing k nearest neighbor points, an insertion sorting algorithm based on queue priority is proposed to reduce the number of judging amount substantially. The experiment results show that GPU parallel computing can improve the LiDAR point cloud filtering efficiency effectively. The acceleration ratio can reach about 9 when data size is above 128 megabytes. [ABSTRACT FROM AUTHOR]

Aim: To analyze the efficacy of surgical resection versus brain biopsy combined with postoperative chemotherapy for primary central nervous system lymphoma (PCNSL) and to discuss a clinically standardized treatment protocol., Material and Methods: Patients with a pathological diagnosis of PCNSL and subsequent chemotherapy between 2016 and 2021 at Northern Jiangsu People?s Hospital were selected and divided into groups according to whether they underwent microsurgical resection or stereotactic needle biopsy. Statistical analyses were performed to compare efficacy and safety in the two groups., Results: A total of 21 patients with PCNSL were identified, of whom 12 underwent resection and 9 underwent diagnostic stereotactic biopsy only. Compared with the resection group, the biopsy group had a higher proportion of deep tumors (55.6% vs. 8.3%, p=0.016), and the mean intraoperative bleeding was significantly reduced (13.33 ± 6.61 mL vs. 170.83 ± 101.04 ml, p < 0.001). In addition, the mean survival time of patients who died during the postoperative follow-up period was shorter (6.83 ± 1.60 vs. 18.56 ± 10.20 months, p=0.016), and the one-year survival rate was lower (33.3% vs. 83.3%, p=0.032). There was no significant difference between the two groups in terms of the mean progression-free survival time or new functional impairment after surgery., Conclusion: For PCNSL, patients who undergo surgical resection have a better outcome than those who undergo biopsy only, suggesting that when the tumor is located at a surgically resectable site, surgical resection should be actively chosen; when the tumor is located at a deep and unresectable site, brain biopsy should be chosen.