Search

Chengxiang Yuan,1,* Siyan Chen,1,* Yiting Ruan,1 Yuntao Liu,1 Haoran Cheng,1 Yaying Zeng,1 Yunbin Chen,1 Qianqian Cheng,2 Guiqian Huang,1 Weilei He,1 Jincai He1 1Department of Neurology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, 325000, Zhejiang, People’s Republic of China; 2School of Mental Health, Wenzhou Medical University, Wenzhou, 325000, Zhejiang, People’s Republic of China*These authors contributed equally to this workCorrespondence: Jincai HeDepartment of Neurology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, 325000, Zhejiang Province, People’s Republic of ChinaTel/Fax +86 0577 55579363Email hjc@wmu.edu.cnBackground: Hemorrhagic transformation (HT) is a severe complication occurring in acute ischemic stroke (AIS) patients. Stress hyperglycemia is frequent in patients with acute illness such as stroke. We aimed to explore the association between stress hyperglycemia and HT in AIS patients.Methods: A total of 287 consecutive participants with HT and 285 age- and sex-matched stroke patients without HT were enrolled in this study. Baseline glucose and glycated hemoglobin (HbA1c) levels were collected to measure stress hyperglycemia. The stress hyperglycemia ratio (SHR) was calculated by dividing the fasting plasma glucose at admission with HbA1c. HT was diagnosed by follow-up imaging assessment, and was radiologically classified as hemorrhagic infarction type (HI) 1 or 2 or parenchymal hematoma type (PH) 1 or 2.Results: Univariate analysis showed that SHR is significantly higher among patients with HT than those without HT. Compared to the patients in the lower three quartiles of SHR, the incidence of HT was significantly higher among patients with the highest quartile of SHR in total population, diabetic and non-diabetic population. We also observed that patients with the highest SHR quartile were associated with an increased risk of hemorrhagic transformation after adjusted for potential covariates (68.4% versus 39.1%; adjusted odds ratio, 2.320; 95% confidence interval, 1.207– 4.459; P=0.012).Conclusion: The stress hyperglycemia ratio, representing the state of stress hyperglycemia, was significantly associated with an increased risk of hemorrhagic transformation in patients with acute ischemic stroke.Keywords: hyperglycemia, hemorrhagic transformation, acute ischemic stroke, stress hyperglycemia

Software comprehension helps us to maintain software better. The developers often use documentation and mutual communication to understand software. However, the incompletion of documentation usually causes some problems such that we cannot acquire the detailed information of software. So researchers are now using key classes as starting points for software understanding and propose several effective methods for key class identification. Such methods are mostly implemented using supervised and unsupervised learning methods; however, they may be subject to class imbalance, insufficient amount of data, and lack of accuracy. In this paper, we propose a novel approach, called COSPA (identifying key Classes in Object-Oriented Software using Preference Aggregation), to identify key classes in software. Our approach is inspired by the social choice theory and uses Kemeny-Young method to find a preference that minimizes the total difference between different pairs of sequences returned by three key class identification methods (i.e., a-index, InDeg, and OSE). Further, we rank the classes in descending order according to that preference set, and a small number of top-ranked classes are considered as key candidates. To better evaluate the performance of our approach, we apply it to six Java software systems and compare it with nine state-of-the-art approaches. In the experiments, we use Friedman to evaluate all methods in the whole set of research subjects; results show that, though no single method works always best in all subject systems, our method indeed performs better than others.

Post-stroke anxiety (PSA) is serious psychosomatic comorbidity among patients with stroke, but whether obesity could be positively associated with PSA is currently unknown. The purpose of this study was to investigate the potential association between obesity and subsequent anxiety risk in patients with stroke. A total of 441 patients with acute ischemic stroke (AIS) onset were consecutively recruited within 7 days, and PSA and post-stroke depression (PSD) were evaluated by using a 14-item Hamilton anxiety scale (HAMA) and 17-item Hamilton depression scale (HAMD) at the end of 1-month follow-up. The odds ratio (OR) with 95% CI was estimated for the incidental PSA by using logistic regression analysis. The incidence of PSA was 25.85% at the end of 1-month follow-up, with a significant difference between patients with and without abdominal obesity. Relative fat mass (RFM) and abdominal obesity were significantly associated with an elevated risk of PSA, and the crude ORs were 1.04 (95% CI: 1.01–1.08) and 1.93 (95% CI: 1.11–3.34), respectively. Even after adjustment for obesity-related risk factors and PSA-related clinical measurements, the association remained to be pronounced with abdominal obesity. However, RFM (OR = 1.03, 95% CI: 0.99–1.06, P = 0.053) and abdominal obesity (OR = 1.31, 95% CI: 0.80–2.15, P = 0.280) were not significantly associated with an elevated risk of PSD. Abdominal obesity was independently associated with the PSA instead of PSD, which may help predict PSA risk in clinical practice. Further prospective clinical studies with a long follow-up duration are warranted to verify this finding.

Machine learning has been pervasively used in a wide range of applications due to its technical breakthroughs in recent years. It has demonstrated significant success in dealing with various complex problems, and shows capabilities close to humans or even beyond humans. However, recent studies show that machine learning models are vulnerable to various attacks, which will compromise the security of the models themselves and the application systems. Moreover, such attacks are stealthy due to the unexplained nature of the deep learning models. In this survey, we systematically analyze the security issues of machine learning, focusing on existing attacks on machine learning systems, corresponding defenses or secure learning techniques, and security evaluation methods. Instead of focusing on one stage or one type of attack, this paper covers all the aspects of machine learning security from the training phase to the test phase. First, the machine learning model in the presence of adversaries is presented, and the reasons why machine learning can be attacked are analyzed. Then, the machine learning security-related issues are classified into five categories: training set poisoning; backdoors in the training set; adversarial example attacks; model theft; recovery of sensitive training data. The threat models, attack approaches, and defense techniques are analyzed systematically. To demonstrate that these threats are real concerns in the physical world, we also reviewed the attacks in real-world conditions. Several suggestions on security evaluations of machine learning systems are also provided. Last, future directions for machine learning security are also presented.

Purpose Post-stroke depression (PSD) is a frequent comorbidity in patients presenting with acute ischemic stroke. Impaired kidney function has been associated with depression in non-stroke subjects. We would like to evaluate whether the estimated glomerular filtration rate (eGFR) on admission is associated with the development of PSD. Patients and methods Total of 268 patients with acute ischemic stroke were recruited and completed 1-month follow-up visit. eGFR was calculated from the serum creatinine value, race, age, and sex by using the chronic kidney disease epidemiology collaboration equation (CKD-EPI creatinine equation). The 17-item Hamilton Depression Scale was used to evaluate depression symptoms. Patients with a depression score of ≥7 were evaluated using the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders, 4th edition, for diagnosing post-stroke depression at 1 month. Meanwhile, 114 normal control subjects were also recruited. Results Ninety-three (34.7%) patients were diagnosed as having PSD at 1 month. There was a significant intergroup difference in eGFR levels within 24 hrs after admission (F=13.608, p

Hemorrhagic transformation (HT) is a severe complication occurring in acute ischemic stroke (AIS) patients. We explored the association between low triiodothyronine (T3) syndrome and HT in AIS patients. A total of 208 consecutive participants with HT and 208 age- and sex-matched stroke patients without HT were enrolled in this study. HT was diagnosed by follow-up imaging assessment, and was radiologically classified as hemorrhagic infarction (HI) type 1 or 2 or parenchymal hematoma (PH) type 1 or 2. HT was also classified into asymptomatic or symptomatic. The incidence of low T3 syndrome was significantly higher among patients who developed HT than among those without HT. Moreover, the more severe the HT, the lower the detected T3 levels. Multivariate-adjusted binary logistic regression showed that low T3 syndrome was an independent risk factor for HT and symptomatic HT in AIS patients. Low T3 syndrome was also significantly associated with a higher risk of PH, but not with the risk of HI. Thus, low T3 syndrome was independently associated with the risk of HT, symptomatic HT, and severe HT (PH) in AIS patients, which suggests monitoring T3 could be a useful means of preventing HT in patients with ischemic stroke.

Image hashing schemes have been widely used in content authentication, image retrieval, and digital forensic. In this paper, a novel image hashing algorithm (SSL) by incorporating the most stable keypoints and local region features is proposed, which is robust against various content-preserving manipulations, even multiple combinatorial manipulations. The proposed algorithm combines S_cale invariant feature transform (SIFT) with S_aliency detection to extract the most stable keypoints. Then, the L_ocal binary pattern (LBP) feature extraction method is exploited to generate local region features based on these keypoints. After that, the information of keypoints and local region features are merged into a hash vector. Finally, a secret key is used to randomize the hash vector, which can prevent attackers from forging the image and the hash value. Experimental results demonstrate that the proposed hashing algorithm can identify visually similar images which are under both single and combinatorial content-preserving manipulations, even multiple combinations of manipulations. It can also identify maliciously forged images which are under various content-changing manipulations. The collision probability between hashes of different images is nearly zero. Besides, the evaluation of key-dependent security shows that the proposed scheme is secure that an attacker cannot forge or estimate the correct hash value without the knowledge of the secret key.

Background and purpose Previous studies have established that vitamin D was associated with stroke. The purpose of this study was to investigate the relationship between vitamin D and 5‐year outcome of patients with stroke including acute ischemic stroke (AIS) and intracranial hemorrhage (ICH) stroke. Methods Serum 25‐hydroxyvitamin D levels were prospectively analyzed in patients admitted to the First Affiliated Hospital of Wenzhou Medical University from 2013 to 2015. Modified Rankin scale (mRS) was used to evaluate their 5‐year functional outcome, and univariate and multivariate logistic regressions were applied to evaluate the effects of vitamin D on stroke outcome. Results In total, 668 patients diagnosed with stroke were recruited, and 420 completed the 5‐year follow‐up. Ninety‐five patients experienced poor outcome in the 5 years since stroke onset. Vitamin D levels in patients with poor outcome showed significant differences compared to good outcome patients (p

Jingjie Hu,* Liuyuan Wang,* Kaili Fan, Wenwei Ren, Qiongzhang Wang, Yiting Ruan, Chengxiang Yuan, Guiqian Huang, Jincai He Department of Neurology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, 325000, Zhejiang, People’s Republic of China*These authors contributed equally to this workCorrespondence: Jincai He; Guiqian HuangDepartment of Neurology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, 325000, Zhejiang Province, People’s Republic of ChinaTel +86 577-555780166; +86 577-555781380Fax +86 577-55578033; +86 577-55578999Email hjc@wmu.edu.cn; huangguiqian123@126.comObjective: Inflammation plays an important role in stroke. Many inflammatory markers in peripheral blood are proved to be associated with stroke severity or prognosis. But few comprehensive models or scales to evaluate the post-stroke depression (PSD) have been reported. In this study, we aimed to compare the level of systemic inflammation markers between PSD and non-PSD patients and explore the association of these inflammatory markers with PSD.Methods: Totally, 432 ischemic stroke patients were consecutively enrolled in the study and received 1 month follow-up. We used the 17-Hamilton Rating Scale to measure depressive symptoms at 1 month after stroke. With the Hamilton Depression Scale score of > 7, patients were diagnosed with PSD. Systemic immune-inflammation index (SII), neutrophil-to-lymphocyte (NLR), platelet-to-lymphocyte (PLR) and derived neutrophil-to-lymphocyte ratio (dNLR) were calculated from the admission blood work.Results: Finally, 129 patients (30.5%) were diagnosed with PSD at 1 month. PSD patients showed significantly higher levels of SII (501.27 (345.43– 782.58) vs 429.60 (315.64– 570.98), P=0.001), NLR (2.36 (1.77– 3.82) vs 2.17 (1.56– 2.80), P=0.010), dNLR (1.67 (1.30– 2.51) vs 1.54 (1.16– 1.99), P=0.009), PLR (124.65 (95.25– 155.15) vs 109.22 (92.38– 142.03), P=0.015), especially SII at admission as compared to non-PSD patients. In the logistic analysis, SII value (> 547.30) was independently associated with the occurrence of PSD (OR=2.181, 95% CI=1.274– 3.732, p =0.004), better than dNLR (OR=1.833, 95% CI=1.071– 3.137, p =0.027), PLR (OR= 1.822, 95% CI=1.063– 3.122, p =0.029) and NLR (OR =1.728, 95% CI=1.009– 2.958, p =0.046).Conclusion: Increased SII, PLR, dNLR, NLR, particularly SII at admission, are significantly correlated with PSD and may add some prognostic clues to find early discovery of PSD.Keywords: inflammatory markers, post-stroke depression, systemic immune-inflammation index, neutrophil-to-lymphocyte, platelet-to-lymphocyte, derived neutrophil-to-lymphocyte ratio

Background Hemorrhagic transformation (HT) is a serious neurological complication of acute ischemic stroke (AIS) after revascularization. The majority of AIS patients do not have atrial fibrillation (AF) which could also develop into HT. In this study, we aimed to explore whether hemostasis parameters are risk factors of HT in non-AF patients. Methods We consecutively enrolled 285 AIS patients with HT. Meanwhile, age- and sex-matched 285 AIS patients without HT were included. The diagnosis of HT was determined by brain CT or MRI during hospitalization. All patients were divided into two subgroups based on the presence of AF and explore the differences between the two subgroups. Blood samples were obtained within 24 h of admission, and all patients were evenly classified into three tertiles according to platelet counts (PLT) levels. Results In this study, we found the first PLT tertile (OR = 3.509, 95%CI = 1.268–9.711, P = 0.016) was independently associated with HT in non-AF patients, taking the third tertile as a reference. Meanwhile, we also found mean platelet volume (MPV) (OR = 0.605, 95%CI = 0.455–0.805, P = 0.001) and fibrinogen (FIB) (OR = 1.928, 95%CI = 1.346–2.760, P P = 0.027) and FIB (OR = 1.298, 95%CI = 1.047–1.610, P = 0.018) were significantly associated with long-term outcomes in non-AF HT patients. Conclusions Low PLT, low MPV, and high FIB levels were independently associated with HT in non-AF patients. Additionally, MPV and FIB levels were significantly associated with unfavorable long-term outcomes in non-AF HT patients. Our study showed that hemostasis functions at admission may be beneficial for clinicians to recognize patients with a high risk of HT at an early stage and improve unfavorable long-term outcomes in non-AF patients.

Introduction Hemorrhagic transformation (HT) is a complex and multifactorial complication among patients with acute ischemic stroke (AIS), and the inflammatory response has been considered as a risk factor for HT. We aimed to evaluate the stratification of FAR (fibrinogen‐to‐albumin ratio), an inflammatory biomarker, in HT patients. Methods A total of 256 consecutive stroke patients with HT and 256 age‐ and gender‐matched stroke patients without HT were included in this study. HT during hospitalization was diagnosed by follow‐up imaging assessment and was classified into hemorrhagic infarction (HI) and parenchymal hematoma (PH) according to the recommendations of European Cooperative Acute Stroke Study II classification. Blood samples were obtained at admission. Results Higher levels of FAR were observed in patients with HT compared with the non‐HT group [10.29 (8.39–12.95) vs. 8.60 (7.25–10.8), p .05]. Patients were assigned to groups of high FAR (≥9.51) and low FAR (, This study is the first to explore the association between fibrinogen‐to‐albumin ratio (FAR) and hemorrhagic transformation (HT) in acute ischemic stroke patients. We observed that FAR was associated with HT. The calculation of FAR may assist clinicians identify patients at elevated risk of HT and make better‐informed individualized treatment.

Background: Hemorrhagic transformation (HT) is a frequent, often asymptomatic event that occurs after acute ischemic stroke (AIS). Liver fibrosis, usually subclinical, is common and crucial in the development of liver disease. We aimed to investigate the association between liver fibrosis and HT in patients with AIS. Methods: We performed a single-center and retrospective study. A total of 185 consecutive participants with HT and 199 age- and sex-matched stroke patients without HT were enrolled in this study. We calculated one validated fibrosis index-Fibrosis-4 (FIB-4) score-to assess the extent of liver fibrosis. HT was detected by routine CT or MRI and was radiologically classified as hemorrhagic infarction type 1 or 2 or parenchymal hematoma type 1 or 2. HT was also classified into asymptomatic or symptomatic. We used logistic regression models adjusted for previously established risk factors to assess the risks for HT. Results: The median FIB-4 score was significantly higher among patients who developed HT than among those without HT, whereas standard hepatic assays were largely normal. Patients were assigned to groups of high FIB-4 score and low FIB-4 score based on the optimal cutoff value. Compared with the subjects in the low-FIB-4-score group, incidence of HT for the high-FIB-4-score group was significantly higher. After adjustment for potential confounders, the patients with high FIB-4 score had 3.461-fold risk of HT in AIS compared to the patients with low FIB-4 score [odds ratio, 3.461 (95% CI, 1.404-8.531)]. Conclusion: Liver fibrosis, measured by FIB-4 score, was independently associated with the risk of HT in AIS patients.

Objective Reduced thiamine (vitamin B1) had been reported to be associated with cognitive impairment caused by Alzheimer disease. Our study is to explore the association between thiamine and cognitive impairment after acute ischemic stroke. Materials and Methods One hundred and eighty two patients with acute cerebral infarction were recruited within the first 24 hr after admission. Thiamine and other vitamin Bs of peripheral blood samples were measured. Patients were divided into with poststroke cognitive impairment (PSCI) and non‐PSCI according to the score of MMSE and the degree of education. Results Reduced thiamine (, PSCI shows a high incidence on acute stage, and more than a half (54.4%) of patients suffered with PSCI. Reduced serum thiamine was associated with early cognitive impairment in patients with acute infarction after adjusting for potential confounding factors. Advanced age, lower education, diabetes mellitus, left hemisphere infarction, and higher NIHSS were also independent risk factors for PSCI.

Objective Withdrawal symptoms are common during methamphetamine (METH) abstinence. This study aimed to explore the association between serum interleukins and withdrawal symptoms during METH abstinence. Methods This study recruited 120 METH users, and 94 of them completed the 2-week follow-up. Serum interleukin-1β, 6,8,10 were tested at admission. Withdrawal symptoms were assessed by the Methamphetamine Withdrawal Questionnaire (MAWQ). Results Serum IL-8 levels were positively correlated with MAWQ scores at the 2-week endpoint (r = .257, p = .013). The variation of the MAWQ scores during the 2-week follow-up was negatively correlated with serum IL-8 levels at admission (r = -.249, p = .026). Serum IL-8 levels remained associated with the severity of METH withdrawal symptoms (β = .363, p = .023), after adjusting for potential confounders. Limitations This study did not include normal controls. Most patients were male and cigarette smokers. Patients were only followed up for 2 weeks, and their toxicology data were not collected. Interleukins were only measured at admission, and were tested in serum, not in the cerebrospinal fluid. Conclusions Our study demonstrated that higher serum IL-8 levels may predict more severe withdrawal symptoms at 2 weeks after METH abstinence.

Recently, the natural language processing- (NLP-) based intelligent question and answer (Q&A) robots have been used ubiquitously. However, the robustness and security of current Q&A robots are still unsatisfactory, e.g., a slight typo in the user’s question may cause the Q&A robot unable to return the correct answer. In this paper, we propose a fast and automatic test dataset generation method for the robustness and security evaluation of current Q&A robots, which can work in black-box scenarios and thus can be applied to a variety of different Q&A robots. Specifically, we propose a dependency parse-based adversarial examples generation (DPAEG) method for Q&A robots. DPAEG first uses the proposed dependency parse-based keywords extraction algorithm to extract keywords from a question. Then, the proposed algorithm generates adversarial words according to the extracted keywords, which include typos and words that are spelled similarly to the keywords. Finally, these adversarial words are used to generate a large number of adversarial questions. The generated adversarial questions which are similar to the original questions do not affect human’s understanding, but the Q&A robots cannot answer these adversarial questions correctly. Moreover, the proposed method works in a black-box scenario, which means it does not need the knowledge of the target Q&A robots. Experiment results show that the generated adversarial examples have a high success rate on two state-of-the-art Q&A robots, DrQA and Google Assistant. In addition, the generated adversarial examples not only affect the correct answer (top-1) returned by DrQA but also affect the top-k candidate answers returned by DrQA. The adversarial examples make the top-k candidate answers contain fewer correct answers and make the correct answers rank lower in the top-k candidate answers. The human evaluation results show that participants with different genders, ages, and mother tongues can understand the meaning of most of the generated adversarial examples, which means that the generated adversarial examples do not affect human’s understanding.

Background: Hemorrhagic transformation (HT) is a complication that may cause neurological deterioration in patients with acute ischemic stroke. Both neutrophil and platelet have been associated with the stroke progression. The aim of this study was to explore the relationship between neutrophil-to-platelet ratio (NPR) and HT after acute ischemic stroke.Methods: A total of 279 stroke patients with HT were consecutively recruited. HT was diagnosed using magnetic resonance imaging (MRI) or computed tomography (CT) and classified into hemorrhagic infarction (HI) and parenchymal hematoma (PH). Blood samples for neutrophil and platelet counts were obtained at admission. Meanwhile, 270 age- and gender-matched controls without HT were included for comparison.Results: Among the patients with HT, 131 patients had PH and 148 patients had HI. NPR was higher in patients with PH than those with HI or non-HT [36.8 (23.7–49.2) vs. 26.6 (17.9–38.3) vs. 19.1 (14.8–24.8), P < 0.001]. After adjustment for potential confounders, high NPR remained independently associated with the increased risk of HT (OR = 2.000, 95% CI: 1.041–3.843, P = 0.037). NPR (>39.9) was independently associated with PH (OR = 2.641, 95% CI: 1.308–5.342, P = 0.007).Conclusions: High NPR was associated with the increased risk of HT especially PH in patients with acute ischemic stroke.

Objectives Several studies have demonstrated that anemia was associated with cognitive impairment. The aim of this study was to explore the relationship between hemoglobin and cognitive impairment in patient with acute ischemic stroke. Methods In total, 326 patients with acute ischemic stroke were followed up for 1 month. The main outcome was the incidence and severity of poststroke cognitive impairment (PSCI) assessed by Mini-Mental State Examination (MMSE). The impact of hemoglobin levels and anemia on PSCI was assessed by multiple regression models controlling for potential confounders. Results During the 1-month follow-up, 193 (59.2%) patients developed PSCI. Anemia was independently associated with PSCI (OR = 3.637; 95% CI, 1.216-10.881; P = .021) after adjusting for demographics, vascular risk factors, stroke severity, and functional outcome. When the hemoglobin levels stratified into tertiles, higher hemoglobin levels were associated with better cognitive function. This result was however not significant after adjusting for the same confounders above. Conclusions Low hemoglobin levels are associated with an increased risk of PSCI. Targeted interventions in this population may reduce the incidence of PSCI and require further evaluation.

Adipogenesis is essential in in vitro experimentation to assess differentiation capability of stem cells, and therefore, its accurate measurement is important. Quantitative analysis of adipogenic levels, however, is challenging and often susceptible to errors due to non-specific reading or manual estimation by observers. To this end, we developed a novel adipocyte quantification algorithm, named Fast Adipogenesis Tracking System (FATS), based on computer vision libraries. The FATS algorithm is versatile and capable of accurately detecting and quantifying percentage of cells undergoing adipogenic and browning differentiation even under difficult conditions such as the presence of large cell clumps or high cell densities. The algorithm was tested on various cell lines including 3T3-L1 cells, adipose-derived mesenchymal stem cells (ASCs), and induced pluripotent stem cell (iPSC)-derived cells. The FATS algorithm is particularly useful for adipogenic measurement of embryoid bodies derived from pluripotent stem cells and was capable of accurately distinguishing adipogenic cells from false-positive stains. We then demonstrate the effectiveness of the FATS algorithm for screening of nuclear receptor ligands that affect adipogenesis in the high-throughput manner. Together, the FATS offer a universal and automated image-based method to quantify adipocyte differentiation of different cell lines in both standard and high-throughput workflows. Electronic supplementary material The online version of this article (10.1186/s13287-019-1141-0) contains supplementary material, which is available to authorized users.

Recently, the natural language processing (NLP) based intelligent question and answering (QA (2) sensitive words are not filtered efficiently and completely; (3) QA (4) unsatisfactory matching algorithms and inappropriate matching threshold settings. Then, we design and implement two types of evaluation tests, bad language and user’s typos, to evaluate the robustness of several state-of-the-art Q&A robots. Experiment results show that these common inputs (bad language and user’s typos) can successfully make these Q&A robots malfunction, denial of service, or speaking dirty words. Besides, we also propose possible countermeasures to enhance the robustness of these Q&A robots. To the best of the authors’ knowledge, this is the first work on analyzing the robustness and security problems of intelligent Q&A robots. This work can hopefully help provide guidelines to design robust and secure Q&A robots.

A sentiment analysis (SA) method for Chinese is presented in this paper. First, natural language processing (NLP) technique is used to analyze the interdependence relationship between components of the sentence, then the syntax components are detected which affect the sentence sentiment. Finally, a sentence sentiment is calculated. The sentiment of some terms is variable. To improve the precision and recall of our method, domain ontology is used to analyze the sentiment of those terms.

k Nearest-Neighbor(kNN) algorithm over sensor network has been extensively studied so far. In order to solve the hotspot problem and prolong the lifetime of wireless sensor networks, the paper proposes a “hotspot cooling” technique in which data of sensor nodes that have little energy can be automatically sent to other sensor nodes in an optimal approach. Extensive experiments indicate that this algorithm can reduce energy consumption of sensor nodes with low power, and efficiently extend the lifetime of wireless sensor networks.

This paper proposes an efficient probabilistic indexing scheme called Probabilistic Multiple-Feature-Tree(PMF-Tree) to facilitate an interactive retrieval of Chinese calligraphic manuscript images based on multiple features such as contour points, character styles and types of character. Different from conventional character retrieval and indexing methods [18] which only adopts shape similarity as a query metric, our proposed indexing algorithm allows user to choose the above three kinds of features they prefer to as query elements. Moreover, a probabilistic modal is introduced to refine the retrieval result. Comprehensive experiments are conducted to testify the effectiveness and efficiency of our proposed retrieval and indexing methods respectively.

Though the utility of domain Ontologies is now widely acknowledged in an increasing number of domains, a critical task of identifying, defining, and entering the concept definitions is still intractable. Nowadays, natural language becomes a more and more important information source, and natural language processing (NLP) becomes more and more practical. To reduce time, cost and making use of NLP, it is highly advisable to refer, in constructing or updating an ontology, to the documents available in the field. In this paper we describe NLTPOnto, a natural language processing based tool is devised to help people to semi-automatically construct ontology.

Evolutionary algorithms EAs are capable to effectively solve hard problems from various fields. When applying EAs to solve specific problems, a huge amount of data will be produced in the process. Due to lack of suitable tools, people seldom explore the rich information in algorithm exclusions, making we are in dark of the dynamics of EAs. Inspired by complex networks research, in this paper, we present a new method to analyse the dynamics of EAs in the form of complex networks. It uses a network model to describe the individuals and their gene relationships. It introduces the parameters of the complex network theory to characterise the evolution of the network. Some properties hidden in the dynamics of EAs are uncovered.