Search

Knowledge Graph-to-Text is a new task in the field of knowledge graph,which aims to transform knowledge graph into readable text describing these knowledge.With the deepening of research in recent years,the generation technology of Graph-to-Text has been applied to the fields of product review generation,recommendation explanation generation,paper abstract generation and so on.The translation model in the existing methods adopts the method of first-plan-then-realization,which fails to dynamically adjust the planning according to the generated text and does not track the static content planning,resulting in incohe-rent semantics before and after the text.In order to improve the semantic coherence of generated text,a Graph-to-Text model based on dynamic memory and two-layer reconstruction enhancement is proposed in this paper.Through three stages of static content planning,dynamic content planning and two-layer reconstruction mechanism,this model makes up for the structural difference between knowledge graph and text,focusing on the content of each triple while generating text.Compared with exis-ting generation models,this model not only compensates for the structural differences between knowledge graphs and texts,but also improves the ability to locate key entities,resulting in stronger factual consistency and semantics in the generated texts.In this paper,experiments are conducted on the WebNLG dataset.The results show that,compared with the current exis-ting models in the task of Graph-to-Text,the proposed model generates more accurate content planning.The logic between the sentences of the generated text is more reasonable and the correlation is stronger.The proposed model outperforms existing methods on me-trics such as BLEU,METEOR,ROUGE,CHRF++,etc.

Object detection is one of the most popular directions in computer vision,which is widely used in military,medical and other important fields.However,most object detection models can only recognize visible objects.There are often covered(invisible) target objects in pictures in daily life.It is difficult for existing object detection models to show ideal detection performance for covered objects in pictures.Therefore,this paper proposes a novel class reasoning model towards covered area in given image based on informed knowledge graph reasoning and multi-agent collaboration(IMG-KGR-MAC).Specifically,first,IMG-KGR-MAC constructs a global prior knowledge graph according to the visible objects of all pictures in a given picture library and the positional relationship between them.At the same time,according to the objects contained in the pictures themselves and their positional relationships,picture knowledge graphs are established for each picture respectively.The covered objects information in each picture is not included in the global prior knowledge graph and the picture's own knowledge graph.Second,deep deterministic policy gradient(DDPG) deep reinforcement learning idea is adopted to build two cooperative agents.Agent 1 selects the “category label” that is most suitable for the covered object from the global prior knowledge graph according to the semantic information of the current picture,and adds it to the knowledge graph of the given picture as a new entity node.Agent 2 further selects 〈entities,relationships〉 from the global prior knowledge graph according to the newly added entities of agent 1,and expands the graph structure associated with the new entity nodes.Third,agent 1 and agent 2 share the task environment and communicate the reward value,and cooperate with each other to carry out forward and reverse reasoning according to the principles of ‘picture covered target(entity) → associated graph structure' and ‘associated graph structure → picture covered object(entity)',so as to effectively estimate the most likely category label of the covered object of a given picture.Experimental results show that,compared with the existing related methods,the proposed IMG-KGR-MAC model can learn the semantic relationship between the covered picture of a given picture and the global prior knowledge graph,effectively overcome the shortcomings of the existing models that it is difficult to detect the covered object,and has good reasoning ability for the covered object.It has more than 20% improvement in many indicators such as MR(mean rank) and mAP(mean average precision).

Automatic text summarization aims to compress a given document, which can efficiently reflect the main idea of the source document with a short summary. At present, abstractive summarization method has become a research hotspot in the field of text summarization because it can paraphrase the source document with flexible and abundant vocabulary. However, existing abstractive summarization model reorganizes original words and adds new words when generating summary. That’s why it can easily cause the inconsistency and low readability. In addition, the traditional supervised learning based on labeled data requires high cost to improve the coherence of summary sentences, which limits the practical application. Therefore, this paper proposes an abstractive text summarization model with coherence reinforcement and no ground truth dependency (ATS_CG). On the one hand, based on the embdding of the source document, the model generates extractive label to describe the filtering process of the key information. And then, the filtered sentence embeddings are decoded by the decoder. On the other hand, based on the original word probability distribution output by the decoder, two types of summarization are generated according to “probability selection” and “Softmax-greedy selection”. And then, the model will compute the overall rewards of the two types of summarization from the two aspects of coherence and content. Next, the model will learn to filter key sentences and decode them through the self-critical policy gradient, so as to generate abstractive summarizaion with high coherence and quality. Experiments show that ATS_CG is superior to the existing text summarization methods in terms of evaluation scores on the whole, even without any ground truth. At the same time, abstractive summarization generated by ATS_CG is also better than the existing methods in coherence, relevance, redundancy, novelty and perplexity.

In recent years, with the rapid development of the Internet, people frequently post comments about a specific object on the Internet. Mastering the critical information from the crowdsourced comments promptly is crucial to the decision-making and service adjustment, with non-negligible application value. Therefore, it is imperative to devote effort to the research on crowdsourced comment integration problem. The goal of the crowdsourced comment integration is to integrate different users comments on the target object into a shorter integrated document by a given compression rate, so as to form a comparatively matched description of the target object according to the public cognition. To solve such problem, a two-phase crowdsourced comment integration method based on reward prediction and policy gradient is proposed. The proposed method does not rely on any man-made ground truth, only requiring the crowdsourced comments. Then, an agent, guided by the experience or reward, will extract key sentence from the crowdsourced comments to generate the integrated comment. Specifically, in the first phase, measuring the content quality of the integrated comment by the relevance and redundancy of sentences, taking the content quality as reward, the long-term reward from selecting a current sentence to the end of the whole comment integration process will be predicted by Q-value, based on which the agent is guided to learn an optimal sentence selection policy. Then, in the second phase, taking the sentiment intensity of the integrated comment as reward, the sentence selection policy learnt in the first phase will be further adjusted by policy gradient, so that the integrated comment generated by the agent can highlight the sentiment intensity from an objective perspective and reflect users attitude more obviously, at the same time, maintaining the content quality. According to the experimental results, compared with the other existing methods, the proposed method can achieve the best overall performance in terms of the content quality as well as the sentiment intensity of the integrated comment, and the time consumed for generation is still controlled at an acceptable level.

In order to provide helpful information for maize variety selection and realizing high yield, the grain filling characteristics of 18 summer maize varieties were investigated under 6/10 (S1), 6/20 (S2), and 6/30 (S3). The results showed that: (1) There were significant differences in the final 100-grain dry weight among varieties and sowing date treatments, characterized by S1 (35.20 g) > S2 (33.45 g) > S3 (31.38 g), and the range of variation was 28.50 g (HM 1) to 36.37 g (JNK 728). (2) The average filling rate (Gave) under different sowing date treatments was S1 (0.74 g 100-grain-1 d-1) > S2 (0.65 g 100-grain-1 d-1) > S3 (0.57 g 100-grain-1 d-1). The average grain filling rate of S1 was 0.09 g 100-grain-1 d-1 and 0.17 g 100-grain-1 d-1 higher than that of S2 and S3, increased by 13.85% and 29.82%, respectively. The Gave of JNK 728 (0.75 g 100-grain-1 d-1) was the highest among varieties, which significantly higher than ZD 958 and XY 335 (0.58 g 100-grain-1 d-1 and 0.67 g 100-grain-1 d-1), increased by 29.31% and 11.94%. The active grain filling period (P) showed that S1 (47.85 d) < S2 (51.39 d) < S3 (55.04 d). (3) The yield under different sowing date treatments was S1 (10,628.67 kg hm-2) > S2 (10,207.65 kg hm-2) > S3 (9144.59 kg hm-2), with S1 4.12% and 16.23% higher than S2 and S3, respectively. Among them, NK 815, MC 121, JNK 729, MC 812, JNK 728 and XY 335 had relatively high yields, with an average of 10,730.56 kg hm-2, significantly higher than ZD 958 (10,080.85 kg hm-2), increased by 6.44%. (4) Correlation analysis showed that yield level was significantly and positively correlated with Gave (0.70**) and grain weight (0.68**), and was significantly and negatively correlated with P (-0.36**); Grain weight was significantly correlated with Gave (0.58 ). In conclusion, maize yield, grain weight and grain filling rate reduced with the sowing date delay, but P increased accordingly, early sowing properly was beneficial to achieve higher yield in Huang-Huai-Hai summer maize region, by selecting varieties such as JNK 728, JNK 729, MC 812, MC 121, NK 815, and XY 335 which characterized by suitable growing period, higher grain filling rate, suitable active grain filling period and higher yield level. JNK 728, characterized by medium-early maturity, higher grain filling rate and yield, was suitable for the later sowing conditions or areas with insufficient accumulated temperature, and could reach physiological maturity in mid-October when sowing at the end of June. [ABSTRACT FROM AUTHOR]

To study the effect of different sowing dates on the process of grain filling characteristics, Jingke 528 (JK528) and Zhengdan 958 (ZD958) were used in the field experiment with two sowing dates (10th April and 14th May). The results showed that the 100-grain dry mass and yield of JK528 and ZD958 with the early spring sowing were significantly higher than that with the spring sowing date, which increased by 6.8% and 10.1%, 17.8% and 9.2%, respectively. The maximum grain filling rate (G max ) and the average grain filling rate (G ave ) were higher with the early spring sowing than that with the spring sowing date, while the active grain filling period (P) showed an opposite trend. The 100-grain mass and yield of JK528 were significantly improved by 7.4 g and 1189.6 kg·hm -2 compared to ZD958, with the increase of 21.6% and 10.8%, respectively. The P of ZD958 was higher than that of JK528, while W max , G max , G ave and T max of JK528 were higher than that of ZD958. The average grain filling rate of JK528 during early and mid grain filling stages was higher than that of late grain filling stages, which was also higher than that of ZD958 at the same sowing date. Furthermore, correlation analysis revealed that the mean temperature and the accumulated temperature were significantly positively correlated with the grain dry matter accumulation. Therefore, the full use of hot-ray resources and the increase of the mean grain-filling rate could improve maize grain yield. The higher grain filling rate during early and mid grain filling stages in JK528 could compensate for the short active grain filling period, leading to higher yields under different sowing dates.

Objective: To explore the level of identification and related factors of depression in physicians of Beijing tertiary general hospitals., Methods: A cross-sectional investigation was used to screen the outpatients, and depression was diagnosed according to Mini International Neuropsychiatric Interview (MINI). Identification of depression was defined as referral to psychiatrist or managed with antidepressant or sedative hypnotics., Results: Among 248 cases of depression, 30 (12.1%) were referred to psychiatrists, 28(8.9%) managed with antidepressant or sedative hypnotics. The identification rate of depression by physicians was 21.0%, and the identification rate by neurologist was 40.4%, which is higher than the rate by cardiologist (22.4%), gastrointestinal physician (3.8%) and gynecologist (2.0%) (χ2=61.077, P<0.05). Visiting neurology department, severity of life impairment, with psychotic distress history, had psychological complaints, co-morbidity with anxiety related to the physician's identification of depression. Among the psychological complaints, only mood disorder and insomnia were related to the physician's identification of depression., Conclusions: The level of depression identification by physician was low, the level of neurologist was higher than that of cardiologist, gastrointestinal physician and gynecologist. The patient with serious life impairment, psychotic distress history, mood disorder and insomnia complaints was easily to be identified.

Objective: To compare the incidence of emotional disorder in patients with acute or stable coronary heart disease., Methods: A total of 298 patients with suspected coronary heart disease (CHD) were designed into three groups based on of coronary angiography results: acute coronary syndrome (ACS, n = 128), stable angina pectoris (SAP, n = 108) and non-CHD (n = 62). All patients were evaluated by Zung Self-rating Depression Scale (SDS), Zung Self-rating Anxiety Scale(SAS) and Hamilton Depression Rating Scale (HRSD) for depression and anxiety before coronary angiography (CAG), 3 days after CAG, and 1 day before discharge., Results: Incidences of depression and anxiety were significantly higher the ACS group (65.6%and 78.9% before CAG; 60.9% and 70.3% 3 days post CAG; 45.3%and 64.8% before discharge) compared patients with SAP (18.5% and 26.9% before CAG; 17.6% and 28.7% 3 days post CAG; 15.7% and 26.9% before discharge, all P < 0.05 vs. ACS) and non-CHD patients (32.3% and 25.8% before CAG; 27.4% and 24.2% 3 days post CAG; 29.0% and 30.6% before discharge, all P < 0.05 vs. ACS) while the depression and anxiety incidences were similar between patients with SAP and non-CHD in this cohort (P > 0.05)., Conclusion: Emotional disorder is common in patients with suspected heart diseases, especially in patients with ACS. Psychological distress of patients with suspected heart disease should be evaluated and treated.

Objective: To explore the prevalence and physician's recognition of depression and anxiety disorder in gastrointestinal out-patients of three tertiary general hospitals in Beijing., Methods: A hospital-based cross-sectional survey was conducted in the gastrointestinal out-patient departments of three tertiary general hospitals in Beijing from May to June 2007. Total 517 subjects were recruited consecutively within a one month period. All the subjects were screened with Hospital Anxiety Depression Scale (HADS). The subjects with HADS score of 8 and over were interviewed and diagnosed by psychiatrists using Mini International Neuropsychiatric Interview (MINI). The physicians made the diagnosis and management without knowing the results of MINI and HADS score., Results: Among the 517 cases, 301 had a HADS score of 8 and above and 244 were interviewed by psychiatrists; the response rate was 81.1%. The prevalence according to Diagnostic and Statistical Manual of Mental Disorders Fourth edition of depressive disorders and anxiety disorders in gastrointestinal outpatients were 15.3%, depressive disorders were 12.0%, anxiety disorders were 6.4%, depression combined anxiety disorder was 3.0%. The prevalence of depression, anxiety, depression combined anxiety were not different between genders (chi2=0.874, chi2=1.797, chi2=0.518, P>0.05) and among different age group (18-34, 35-54, >or=55 years old) (chi2=1.084, chi2=2.735, chi2=0.350, P>0.05). Gastritis and gastrointestinal dysfunction were the major diagnoses in patients with depression and/or anxiety disorders, the rates were 30.6% and 26.4% respectively. The rate of identification of depression and anxiety disorder by physicians was 2.8%., Conclusion: Gastrointestinal out-patients have a high prevalence of depression and anxiety disorder and the rate of identification by physicians was very low.

Objective: To explore the current condition of psychological problems of patients in cardiology clinic., Methods: A total of 347 patients in the cardiology clinic of Prof. HU Da-yi were screened by self-rating depression scale (SDS), Self-rating anxiety scale (SAS), Hamilton rating scale for depression (HRSD) and Hamilton anxiety scale (HAMA)., Results: The rates of patients with somatic disease only or with psychological disorder only were 52.4% and 12.7%, the rate of somatic disease accompanied with psychological disorder was 27.7%. For the patients with psychological problems, there were 31 cases with depression (8.9%), 62 cases with anxiety (17.9%), and 28 cases (8.1%) suffering from mixed depression and anxiety the diagnosis of cardiovascular disease, functional disease or digestive disease was always used by doctors for the patients with psychological problems only, as they always did not accept the diagnosis of psychological disorder., Conclusions: It is very common that patients in cardiology clinic department suffered from psychological problems. The patients with psychological disorder only were easily misdiagnosed as somatic disease especially cardiovascular disease.

Objective: To detect plasma concentrations of vascular endothelial cell growth factor (VEGF) and tissue factor (TF) in children with acute lymphoblastic leukemia (ALL) and explore their clinical significance in ALL., Methods: Thirty-three children with newly diagnosed ALL, including 18 cases of low risk, 7 cases of moderate risk and 8 cases of high risk, were enrolled in this study. Twenty-five patients received a complete remission and 8 cases were in non-remission after conventional remission induction chemotherapy. Plasma concentrations of VEGF and TF in the patients were detected using ELISA before and after treatment. Sixteen healthy children served as normal control group., Results: Plasma concentrations of VEGF and TF in ALL patients before treatment were significantly higher than those in normal controls (P < 0.01). Plasma concentrations of VEGF and TF in the non-remission group before treatment were significantly higher than those in the remission group (P < 0.05) and the control group (P < 0.01). After treatment the plasma concentrations of VEGF and TF in the non-remission group were not significantly reduced and higher than those in the remission and the control groups (P < 0.01). There were significant differences in plasma concentrations of VEGF and TF among the low-risk, moderate-risk and high-risk groups before and after treatment (P < 0.05). Plasma concentrations of VEGF and TF in the high risk group were not significantly reduced after treatment and higher than those in the control group (P < 0.01). A linear correlation was noted between plasma VEGF and TF concentrations in ALL patients before treatment (r=0.50, P < 0.01)., Conclusions: VEGF and TF play an important role in the development of ALL and may be useful to the evaluation of the severity and the outcome in ALL.

Linkage disequilibrium (LD) is one of the most recently focused interests in the field of plant genomics. LD mapping is an effective approach to discovering novel genes and a bridge for connecting structural genomics to phenomics. LD mapping was first applied in plants in 2001. Since then, researches on the structure and extent of LD and LD mapping have been reported in a wide range of plant species. The basic theory of LD and its application in LD mapping, haplotype diversity analysis, htSNP identification and population genetics were reviewed in this paper. And advances of LD research in plants including influences of population structure, gene conversion, epistasis and GxE interactions, and future prospects were also presented. China has abundant germplasm resources, but gene discovery lags behind. Intensive researches on LD will certainly accelerate rapid development of plant genomics, especially the progress of gene discovery based on germplasm resources in China.

Objective: To evaluate the levels of bone morphogenetic protein-15 (BMP-15) in human follicular fluid (FF) and its association with response to ovarian stimulation., Methods: Western blotting was performed to determine the levels of BMP-15 in FF obtained from follicle aspirates in 70 patients undergoing IVF treatment. According to the response to ovarian stimulation the patients were divided into poor responder group and normal responder group., Result: BMP-15 levels in FF of poor responders were significantly higher than those in normal responders (1.01 +/- 0.34 vs 0.77 +/- 0.24, P<0.01)., Conclusion: Increased levels of BMP-15 in FF may be associated with poor response to ovarian stimulation.

Objective: To investigate the possible roles of cyclin B1 and cyclin-dependent kanase1 cyclin-dependent kanase1 play in the pathogenesis and development of endometriosis and their association with the ovarian hormones., Methods: Twenty-nine specimens of ectopic endometrium tissues and 20 specimens of eutopic endometrium tissues were obtained from 29 patients with endometriosis, aged 24 approximately 46. Thirty specimens of endometrium form 30 women without endometriosis were used as controls. The intracellular location of cyclin B1 and Cdc2 was detected by microscopy. Western blotting, RT-PCR, and immunohistochemistry were employed to examine the mRNA expression and protein expression of cyclin B1 and Cdc2. Serum estrogen and progestogen were detected., Results: The expression level of cyclin B1 in the ectopic endometrium from the women with endometriosis was significantly higher than that in the ectopic endometrium tissues from the women with and without endometriosis (both P < 0.05). No significant difference was found between the expression level of cyclin B1 in the ectopic endometrium tissues from the women with and without endometriosis (both P > 0.05). The Cdc2 expression levels were not significantly different among th3 3 groups and the proliferative and secretary stage of endometrium (all P > 0.05). Cyclin B1 expression level was positively correlated with the serum estrogen level, and negatively correlated with the serum progestogen level, and Cdc2 expression was not correlated with the serum sex hormone levels., Conclusion: The expression of cyclin B1 in the ectopic endometrium is higher than normal. Cyclin B1 may be involved in the proliferation of endometrium in endometriosis.

Objective: To investigate whether the limited ovarian stimulation (LOS) could avoid the severe ovarian hyperstimulation syndrome (OHSS) in patients with high risk of OHSS., Methods: Ten infertile patients were diagnosed as polycystic ovarian syndrome (PCOS) with high risk of OHSS. All of them were arranged LOS cycles, a long-induction protocol using gonadotropin releasing hormone agonist (GnRHa) for controlled ovarian hyperstimulation. A full dose of human chorionic gonadotropin (10,000 U) was administered when the leading follicle reached a mean diameter of 12.0 - 14.5 mm. Oocytes were retrieved 36 hours later, followed by conventional in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI) and embryo transfer., Results: Ten patients produced 11 - 35 oocytes, including 4 - 28 oocytes of metaphase II. Embryos were transferred in all of ten patients and eight clinical pregnancies were diagnosed. Unfortunately, ectopic pregnancy occurred in one of the patients. None of them experienced the symptoms of OHSS., Conclusions: LOS may play a valuable role in preventing OHSS and achieving enough mature oocytes for the patients of PCOS. Furthermore, this protocol would not reduce the pregnancy rate of the patients.

Objective: To examine the expression of aquaporin 2 (AQP2) in human endometrium., Methods: Specimens of human endometrium were collected from 87 women at different menstrual cycles, aged 30 +/- 3, 23 cases in the proliferative phase, 30 cases in the early secretory phase and 34 cases in the mid-secretory phase respectively. Immunohistochemistry and Western blotting were utilized to detect the localization and expression of AQP2 in the endometrium. Reverse transcription-polymerase chain reaction (RT-PCR) was employed to detect its messenger RNA. The PCR product was cloned and sequenced., Results: Positive immunoreactivity of AQP2 was found in the epithelia cells and glandular epithelial cells of all specimens of human endometrium at different phases of the menstrual cycle, and all stromal cells were not stained. The reactive substance was primarily distributed in the membrane and cytoplasm, but not in the nuclei of all positive cells. In Western blotting showed dominant bands with relative molecular weight between 35,000 Da and 50,000 Da, which was corresponded to the glycosylated form of AQP2 by the positive control from rat kidney. Semi-quantitative analysis showed that the relative expression of AQP2 in the mid-secretory phase was 1.63 +/- 0.15, significantly higher than those in the early secretory phase (1.33 +/- 0.14, P < 0.05) and that in the proliferative phase (1.03 +/- 0.10, P < 0.01). Message RNA was found out in all cases by RT-PCR and the PCR product was confirmed in nearly exact (99%) consistency with the GenBank by sequencing. AQP2 mRNA was expressed in all normal endometrium at different phases, and was weakly expressed in the endometrium at the proliferative phase. Sequencing showed that the AQP2 sequence was 99% homologous with that in the GenBank., Conclusion: AQP2 expression in the human endometrium suggests that AQP2 may be involved in the regulation of uterine fluid homeostasis influenced by ovarian steroid hormones in the menstrual cycle.

Objective: To investigate the expressions of peripheral lymph node addressin(PNAd) and GlcNAc-6-sulfotransferase (GlcNAc6ST) in endometrium and their impacts on implantation., Methods: PNAd expression in endometrium was examined by immunohistochemistry and Western Blot from 75 women (12 from healthy women, in proliferative phase; 63 from sterile women, of whom, 27 were in early-secretory and 36 in mid-secretory phase). GlcNAc6ST mRNA was examined by real-time PCR in 41 sterile women. The 63 sterile women had underwent IVF-ET and were consequently divided into clinical pregnant (29 cases) and nonpregnant (34 cases) groups., Results: (1) PNAd localized to the membrane and cytoplasm of luminal and glandular epithelia. Staining was patchy and much less intense during the proliferative phase than during the secretory phase. In Western Blot of PNAd, four bands appeared, which were Sgp200, CD34, MAdCAM-1, GlyCAM-1 respectively, and each was positively correlated with the others significantly. The former three molecular levels were significantly higher during the secretory phase as compared with the proliferative phase. Message RNA of GlcNAc6ST was positive in all cases and showed no correlation with any component of PNAd. (2)The expressions of CD34 and GlyCAM-1, but not Sgp200 and MAdCAM-1, were significantly higher in pregnant women than in nonpregnant ones. However, the GlcNAc6ST mRNA level did not differ between groups. (3) No significant difference was found in female age, methods of fertilization, thickness of endometrium on day hCG, cumulative embryo score(CES) and mean score of transferred embryo(MSTE) between the groups., Conclusion: PNAd expression in the human endometrium fluctuates with the menstrual cycle. Elevated CD34 and GlyCAM-1 during the secretory phase might be stimulative factors for embryo implantation. Defect in PNAd expression may account for a portion of unexplained infertility.