Search

Dialogue segmentation is a crucial task for dialogue systems allowing a better understanding of conversational texts. Despite recent progress in unsupervised dialogue segmentation methods, their performances are limited by the lack of explicit supervised signals for training. Furthermore, the precise definition of segmentation points in conversations still remains as a challenging problem, increasing the difficulty of collecting manual annotations. In this paper, we provide a feasible definition of dialogue segmentation points with the help of document-grounded dialogues and release a large-scale supervised dataset called SuperDialseg, containing 9K dialogues based on two prevalent document-grounded dialogue corpora, and also inherit their useful dialogue-related annotations. Moreover, we propose two models to exploit the dialogue characteristics, achieving state-of-the-art performance on SuperDialseg and showing good generalization ability on the out-of-domain datasets. Additionally, we provide a benchmark including 20 models across four categories for the dialogue segmentation task with several proper evaluation metrics. Based on the analysis of the empirical studies, we also provide some insights for the task of dialogue segmentation. We believe our work is an important step forward in the field of dialogue segmentation., Datasets and codes are available at https://github.com/Coldog2333/SuperDialseg

Objective To investigate the expression of PD-1 and PD-L1/2 in T cell subsets and myeloma cells in the bone marrow from newly diagnosed multiple myeloma (NDMM) patients and their relation with clinical features. Methods We collected the bone marrow and clinical data of 22 NDMM patients and 18 cases of healthy controls. We sorted CD4+T cells, CD8+T cells and myeloma cells by flow cytometry, and observed the expression of PD-1 and PD-L1/2. Results Compared with the control group, the proportion of CD8+T cells in the NDMM group was significantly higher, while the ratio of CD4+/CD8+ was significantly lower (both P < 0.05). The expression levels of PD-1 and PD-L2 in CD4+T cells in the NDMM group were significantly higher than those in the control group (both P < 0.05). The expression levels of PD-1, PD-L1 and PD-L2 in T cell subsets and myeloma cells of NDMM patients were not correlated with the gender, age, immune typing, Durie-Salmon stage and subtypes, ISS stage or mSMART3.0 stratification (both P > 0.05). Conclusion Most of MM patients suffering immune abnormality, which may be associated with the mutual immunosuppressive effects between T lymphocytes and plasma cells which expressing PD-1 and PD-L1/2.

Supplementary document

Supporting Information

Objective To investigate the detection and clinical significance of thrombus molecular markers in diffuse large B-cell lymphoma (DLBCL). Methods We collected the blood specimens of 60 patients with DLBCL, involving 23 cases in the initial treatment group, 24 cases in the remission group and 13 cases in the non-remission group, 23 cases in the thrombus group and 37 cases in the non-thrombus group. We selected 46 healthy people in the same period as the control group. The levels of thrombomodulin (TM), plasmin-α2 plasmin inhibitor complex (PIC), tissue plasminogen activator-plasminogen activator inhibitor-1 complex (t-PAIC) and thrombin-antithrombin Ⅲ complex (TAT) in plasma were detected by chemical immunoassay, and the levels of lactate dehydrogenase (LDH) in serum was detected by automatic biochemical analyzer. We analyzed the differences of thrombus molecular markers among groups and prognostic factors. Results The levels of TM and PIC in plasma of lymphoma patients were higher than those in health control group (P < 0.05). The levels of TM and PIC in the initial treatment and non-remission groups were significantly higher than those in the remission group (P < 0.05). The levels of TM, PIC and TAT in thrombus group were higher than those in non-thrombus group (P < 0.05). TM and PIC levels in plasma were closely related to the prognosis of DLBCL patients. PIC was an independent prognostic factor (P < 0.001). TM and PIC levels were correlated with LDH prognostic indicators in lymphoma patients. Conclusion TM and PIC levels in plasma are significantly increased in DLBCL patients. They are expected to be the indicators for effectiveness and prognosis of DLBCL patients.

The association between socioeconomic status (SES) and dietary diversity score (DDS) has been widely discussed, but little is known about the age and cohort effects on DDS and how the SES effect on DDS varies with age and across successive cohorts among urban and rural older adults in China. Thus, this study aimed to examine the temporal change in DDS among Chinese older adults and SES heterogeneities in such change from the perspective of urban–rural differentiation. Data from the Chinese Longitudinal Healthy Longevity Survey (CLHLS) between 2002 and 2018 were used, and a total of 13,728 participants aged between 65 and 105 years were included in this study. A total of eight food groups were used to assess DDS, while education, family income, and perceived income status were used to assess SES. A linear mixed model was used to estimate the age and cohort effects on DDS and their urban–rural and SES disparities. The results show that higher SES, including more education, family income, and perceived income status, was associated with higher DDS (for urban older adults, β = 0.1645, p = 0.0003, β = 0.2638, p < 0.0001, β = 0.2917, p < 0.0001, respectively; for rural older adults, β = 0.0932, p = 0.0080, β = 0.4063, p < 0.0001, β = 0.2921, p < 0.0001, respectively). The DDS of older adults increased with age and across successive cohorts in both urban and rural China. Moreover, we found the three-way interaction effect of SES, age, and cohort was statistically significant in both urban and rural China. Thus, living in an urban area and having higher SES are associated with higher DDS, but these associations change with age and across successive cohorts. The dietary health of earlier cohorts and rural oldest-old in China deserves more attention.

In order to encourage residents to go to primary care facilities, China has set up differentiated basic medical insurance reimbursement ratios. The study aims to use the dynamic point of view of longitudinal data to examine the changes in the impact of basic medical insurance on primary care. The data for this study comes from the Chinese Family Panel Study (CFPS) in 2010, 2012, 2014, 2016, and 2018. We adopted Hierarchal Age-period-cohort-Cross-Classified Random Effects Models (HAPC-CCREM) to examine the changes in the impact of basic medical insurance on primary care. Compared with non-insured groups, participants of the New Rural Cooperative Medical System (coefficient = 0.730) have a relatively high incidence of primary care seeks, while Urban Residents' Basic Medical Insurance (coefficient = −0.482) and Urban Employees' Basic Medical Insurance (coefficient = −0.663) are lower, respectively. Age, period over time and cohort have a more obvious moderating effect on primary care seeks. The study of primary care behavior is an important direction for the construction of a hierarchical medical system. As basic medical insurance is the source of power for the hierarchical medical system, we can provide certain direction for policy formulation on the changes of basic medical insurance in primary care behavior.

In this paper, we introduce the task of learning unsupervised dialogue embeddings. Trivial approaches such as combining pre-trained word or sentence embeddings and encoding through pre-trained language models (PLMs) have been shown to be feasible for this task. However, these approaches typically ignore the conversational interactions between interlocutors, resulting in poor performance. To address this issue, we proposed a self-guided contrastive learning approach named dial2vec. Dial2vec considers a dialogue as an information exchange process. It captures the conversational interaction patterns between interlocutors and leverages them to guide the learning of the embeddings corresponding to each interlocutor. The dialogue embedding is obtained by an aggregation of the embeddings from all interlocutors. To verify our approach, we establish a comprehensive benchmark consisting of six widely-used dialogue datasets. We consider three evaluation tasks: domain categorization, semantic relatedness, and dialogue retrieval. Dial2vec achieves on average 8.7, 9.0, and 13.8 points absolute improvements in terms of purity, Spearman's correlation, and mean average precision (MAP) over the strongest baseline on the three tasks respectively. Further analysis shows that dial2vec obtains informative and discriminative embeddings for both interlocutors under the guidance of the conversational interactions and achieves the best performance when aggregating them through the interlocutor-level pooling strategy. All codes and data are publicly available at https://github.com/AlibabaResearch/DAMO-ConvAI/tree/main/dial2vec., Comment: Accepted as Long Paper at "EMNLP,2022"

Additional file 1: Figure S1. WGS results in HepG2. Figure S2. Representative Sanger sequencing results of gRNA-expression vectors and their targeting sites in HepG2 and Hep3B. Figure S3. The "i-CRISPR" strategy could induce DSBs and cell death in HepG2 cells. Figure S4. Targeted induction of DSB resulted in apoptosis in cancer cells carrying specific mutations. Figure S5. WGS results in DU145. Figure S6. The "i-CRISPR" strategy could suppress DU145 cell viability. Figure S7. The "i-CRISPR" strategy could induce cell death in DU145 cancer cells but not normal 293 T cells. Figure S8. CRISPR targeting new targets could overcome the possible drug resistance problems of our strategy. Figure S9. The "i-CRISPR" strategy could only specifically suppress HCC organoid with the designed mutations. Figure S10. The "i-CRISPR" strategy could not significantly affect the weight, blood routine and other biochemical indicators of PCa PDX mice. Figure S11. Sequences of off-target sites identified by GUIDE-seq for the mixture of Cas9 and 3 gRNAs targeting mutated sites and treated DU145 cells. Figure S12. Quantitative phos-phoproteomics analysis elucidate the molecular mechanism of the "i-CRISPR" strategy in HepG2 cells. Figure S13. Differentially phosphorylated proteins analysis after the "i-CRISPR" strategy treatment. Figure S14. Enrichment of GO terms for differentially phosphorylated proteins among the 3 groups. Figure S15. Activation of autophagy may play critical role in the molecular mechanism for the increased cell death of our strategy-part 1. Figure S16. Activation of autophagy may play critical role in the molecular mechanism for the increased cell death of our strategy-part 2. Figure S17. DNA methylation analysis in control and treated DU145 cells. Figure S18. DNA methylation analysis in control and treated HepG2 cells. Figure S19. Joint analysis of the DNA methylation data in DU145 and HepG2 cells. Figure S20. Gene body region methylation of representative DMR located 312 genes in DU145 and HepG2 cells. Figure S21. Whole-genome sequencing results of DU145 and the 3 DU145 derived single cells (A6, B12, B13) cultured in our laboratory. Supplementary materials and methods.

Coherent optical fiber systems can achieve long-distance, large-capacity and high data-rate transmissions. The system performance of communication systems is generally evaluated with regard to the data capacity and the transmission reach. In this work, the performance of multi-channel (up to C-band) Nyquist-spaced coherent optical communication systems has been assessed in terms of achievable information rates, transmission distances and signal-to-noise ratios, considering different influencing factors, such as nonlinearity compensation, signal input power and modulation format. Numerical simulations and enhanced Gaussian noise (EGN) model have been carried out for different modulation formats including quadrature phase shift keying (QPSK), 16-ary quadrature amplitude modulation (16-QAM), 64-QAM and 256-QAM. It is found that in C-band (151-channel) Nyquist-spaced systems, the achievable information rates at the transmission distance of 6000 km are 19.3 Tbit/s for dual-polarization QPSK (DP-QPSK), 30.9 Tbit/s for DP-16QAM, 32.0 Tbit/s for DP64QAM and 32.2 Tbit/s for DP-256QAM, respectively, when electronic dispersion compensation is applied only. Such achievable information rates can be increased up to 38.3 Tbit/s for DP-16QAM, 47.2 Tbit/s for DP-64QAM and 47.8 Tbit/s for DP-256QAM, respectively, when the nonlinearity compensation is employed.\ud

This paper introduces an unsupervised loss for training parametric deformation shape generators. The key idea is to enforce the preservation of local rigidity among the generated shapes. Our approach builds on an approximation of the as-rigid-as possible (or ARAP) deformation energy. We show how to develop the unsupervised loss via a spectral decomposition of the Hessian of the ARAP energy. Our loss nicely decouples pose and shape variations through a robust norm. The loss admits simple closed-form expressions. It is easy to train and can be plugged into any standard generation models, e.g., variational auto-encoder (VAE) and auto-decoder (AD). Experimental results show that our approach outperforms existing shape generation approaches considerably on public benchmark datasets of various shape categories such as human, animal and bone.

Severe blockage in myeloid differentiation is the hallmark of acute myeloid leukemia (AML). Trdmt1 plays an important role in hematopoiesis. However, little is known about the function of Trdmt1 in AML cell differentiation. In the present study, Trdmt1 was up-regulated and miR-181a was down-regulated significantly during human leukemia HL-60 cell differentiation after TAT-CT3 fusion protein treatment. Accordingly, miR-181a overexpression in HL-60 cells inhibited granulocytic maturation. In addition, our "rescue" assay demonstrated that Trdmt1 3'-untranslated region promoted myeloid differentiation of HL-60 cells by sequestering miR-181a and up-regulating C/EBPα (a critical factor for normal myelopoiesis) via its competing endogenous RNA (ceRNA) activity on miR-181a. These findings revealed an unrecognized role of Trdmt1 as a potential ceRNA for therapeutic targets in AML.

Due to the specific cofactor dependence and the potent catalytic capability, DNAzymes have shown great promise in the application of biosensors. However, it poses significant challenges for researchers to enhance the sensitivities of DNAzyme-based biosensing platforms. In this work, an ultra-sensitive, label-free, and quick-responsive DNAzyme biosensor is developed based on whispering gallery mode (WGM) optofluidic resonator with a hybrid amplification from liquid crystals (LCs) and Au nanoparticles (AuNPs). When a DNAzyme is cleaved by biotargets, the event of DNA hybridization is triggered and will result in the orientation transition of LC molecules (in the resonator core). According to the quadruple amplification from the disturbance of AuNPs, the excess polarizability of adhering biomolecules, the orientation variation of LC molecules, and the WGM resonance, the spectral wavelength shift can be employed as the sensing parameter to indicate the information of biotarget. L-histidine molecules are applied as the analyte. An ultra-low detection limit at a sub-femtomole level (~ 5×10-16 M) has been achieved using this highly-selective biosensor, which is eight orders of magnitude lower compared to the reported LC-based L-histidine sensor. In addition to L-histidine, this proposed optofluidic scheme with LC-AuNP hybrid amplification can also extend the applicability of DNAzyme-based biosensors for the detection of various molecules.\ud \ud

Recently, open domain multi-turn chatbots have attracted much interest from lots of researchers in both academia and industry. The dominant retrieval-based methods use context-response matching mechanisms for multi-turn response selection. Specifically, the state-of-the-art methods perform the context-response matching by word or segment similarity. However, these models lack a full exploitation of the sentence-level semantic information, and make simple mistakes that humans can easily avoid. In this work, we propose a matching network, called sequential sentence matching network (S2M), to use the sentence-level semantic information to address the problem. Firstly and most importantly, we find that by using the sentence-level semantic information, the network successfully addresses the problem and gets a significant improvement on matching, resulting in a state-of-the-art performance. Furthermore, we integrate the sentence matching we introduced here and the usual word similarity matching reported in the current literature, to match at different semantic levels. Experiments on three public data sets show that such integration further improves the model performance., 10 pages, 4 figures

A high precision demodulation method in frequency domain is proposed for a polarized low-coherence interferometer with location-dependent birefringence dispersion. By minimizing the frequency nonlinearity caused by dispersion effects, the proposed method avoids the jump errors of traditional frequency domain analysis method and exactly retrieve the absolute phase at multiselected-wavenumber point simultaneously, which makes it superior in accuracy, stability, and measurement range. We carried out experiments with an optical fiber Fabry–Perot pressure sensing system to verify the effectiveness of the proposed method. The experiment result showed that the measured error was less than 0.049 kPa and the measurement range was widened to 285 kPa, which could be wider with suited experimental equipment and is no limit in theory.

To investigate how the changes in performance and the microbial community of the co-digestion system of Pennisetum hybrid and pig manure, two co-digestion systems in a semi-continuous mode were established at different grass:manure mixture ratios (50:50 and 75:25), and at variable organic loading rates (OLRs). The two reactors were in a steady-state at the OLRs of 2.0–5.0 g VS/(L·d), with the specific and volumetric biogas yields of 383.86 ± 65.13 to 574.28 ± 72.04 mL/g VS and 0.87 ± 0.07 to 2.36 ± 0.13 m3/(m3·d), respectively. The co-digestion system with a mixture ratio of 75:25 failed at an OLR of 5.5 g VS/(L⋅d). This failure could be attributed to the accumulation of volatile fatty acids (VFAs) owing to the imbalance between acid-production and -oxidation bacteria. By contrast, the co-digestion system with mixture ratio of 50:50 failed at an OLR of 7.0 g VS/(L⋅d), which was likely due to mechanical issues or improper reactor configuration. The genus Proteiniphilum contributed to the increase in total ammonia nitrogen. These findings provide useful guidance for optimizing co-digestion system, enhancing reactor performance and improving the wastes treatment.

Background The link between prostate cancer (PCa) development and aberrant expression of genes located on the Y chromosome remains unclear. Objective To identify Y-chromosomal long noncoding RNAs (lncRNAs) with critical roles in PCa and to clarify the corresponding mechanisms. Design, setting, and participants Aberrantly expressed lncRNAs on the Y chromosome were identified using transcriptome analysis of PCa clinical samples and cell lines. Biological functions and molecular mechanisms of the lncRNAs were revealed using in vitro and in vivo experimental methods. Outcome measurements and statistical analysis Experiments and outcome measurements were performed in duplicate or triplicate. Wilcoxon signed-rank test was employed for comparison of RNA levels in clinical cohorts. Analysis of variance was employed for comparisons among multiple groups. Results and limitations In most patients with PCa, TTTY15 was the most elevated lncRNA located on the Y chromosome. Knockout of this lncRNA by two different CRISPR-Cas9 strategies suppressed PCa cell growth both in vitro and in vivo. TTTY15 promoted PCa by sponging the microRNA let-7, consequently increasing CDK6 and FN1 expression. FOXA1 is an upstream regulatory factor of TTTY15 transcription. Conclusions The Y-chromosomal lncRNA TTTY15 was upregulated in most PCa tissues and could promote PCa progression by sponging let-7. Patient summary We found that TTTY15 levels were frequently elevated in prostate cancer (PCa) tissues compared with those in paracancerous normal tissues in a large group of PCa patients, and we observed a tumour suppressive effect after TTTY15 knockout using CRISPR/Cas9. These results may have therapeutic implications for PCa patients.

Coherent Anti-Stokes Raman Scattering (CARS) microscopy has attracted lots of attention because of the advantages, such as noninvasive, label-free, chemical specificity, intrinsic three-dimension spatial resolution and so on. However, the temporal overlap of pump and Stokes has not been solved owing to the ultrafast optical pulse used in CARS microscopy. We combine interference spectrum of residual pump in Stokes path and nonlinear Schrodinger equation (NLSE) to realize the temporal overlap of pump pulse and Stokes pulse. At first, based on the interference spectrum of pump pulse and residual pump in Stokes path, the optical delay is defined when optical path difference between pump path and Stokes path is zero. Then the relative optical delay between Stokes pulse and residual pump in PCF can be calculated by NLSE. According to the spectrum interference and NLSE, temporal overlap of pump pulse and Stokes pulse will be realized easily and the imaging speed will be improved in CARS microscopy.

In this paper, we design an integrated algorithm to evaluate the sentiment of Chinese market. Firstly, with the help of the web browser automation, we crawl a lot of news and comments from several influential financial websites automatically. Secondly, we use techniques of Natural Language Processing(NLP) under Chinese context, including tokenization, Word2vec word embedding and semantic database WordNet, to compute Senti-scores of these news and comments, and then construct the sentimental factor. Here, we build a finance-specific sentimental lexicon so that the sentimental factor can reflect the sentiment of financial market but not the general sentiments as happiness, sadness, etc. Thirdly, we also implement an adjustment of the standard sentimental factor. Our experimental performance shows that there is a significant correlation between our standard sentimental factor and the Chinese market, and the adjusted factor is even more informative, having a stronger correlation with the Chinese market. Therefore, our sentimental factors can be important references when making investment decisions. Especially during the Chinese market crash in 2015, the Pearson correlation coefficient of adjusted sentimental factor with SSE is 0.5844, which suggests that our model can provide a solid guidance, especially in the special period when the market is influenced greatly by public sentiment.

This paper proposes a novel method for the creation of user-defined freeform feature (UDFFF) from existing surface model. The developed methodology has four major steps: selecting the region of interest (ROI), obtaining the characteristic curves from the surface model, reconstructing the ROI surface, and parameterizing the UDFFF. To quickly create UDFFF on the ROI, the representation of freeform feature based on characteristic curves is proposed. First, the characteristic curves of ROI are obtained automatically or by a small amount of human–computer interaction. In particular, user can directly and easily draw the curves on surface to satisfy the designer's requirements. Second, from the set of characteristic curves that is obtained, freeform surface is reconstructed by a suitable surface generation method, e.g., sweeping, revolving, skinning, and filling. Finally, parameterization of the freeform feature is configured, which primarily involves establishing parameters and building the mappings between feature parameters and shape data. Moreover, according to the degrees of freedom, two types of parameterization of freeform features may be described: one (low degrees of freedom) can be accurately described by the characteristic curves and the other (high degrees of freedom) can be approximated with these curves. The proposed UDFFF creation method has been tested with examples based on the surface models. Experience with our prototype system indicates that it allows untrained users to create UDFFFs from the surface models.

Detecting global exact symmetries in CAD models is of great importance in the research of CAD/CAE integration. Therefore, a method is proposed in this paper to rapidly detect the global exact rotational and reflectional symmetries in feature-based CAD models. The theories of determining the symmetries of the Boolean combinations of the features are framed. Based on these theories, our approach is processed as follows. First, the features of the CAD models are classified into congruent feature sets. Next, through the study on the relationship between feature information and the symmetries of features, by using only feature information, as many symmetries of the feature sets as possible are detected. Then these feature sets are sorted into an ordered sequence. Finally, symmetries of the entire model can be derived by successively merging and verifying the symmetries of feature sets in the ordered sequence. Experimental results show that the global exact symmetries can be robustly and rapidly detected.

We present a Frequency-domain optical coherence tomography (FD-OCT) system for intravascular imaging. The whole system consists of 4 parts: imaging catheter (IC), motion controller (MC), optical engine (OE) and data acquisition and processing module (DAPM). By using swept laser source with 50 KHz sweep speed and high speed optical rotary jiont, we can get 100 frames per second (fps) and scan 20 mm vessel per second.

Through analysing the spectrum of the cascade optical fiber Fabry-Perot pressure sensor composed with three reflecting surfaces,the lower envolope of the spectrum can be used to demodulate the sensor′s cavity length.A novel algorithm based on fitting of interference-order is proposed.Simulation analysis is carried out with the cavity length of F-P2 variating between 160 μm to 215 μm.It is shown that the results precision can be improved using the novel method at the different signal to noise ratio(SNR),and is not impacted by sensor′s cavity length and the optical spectrum resolution ratio at the same SNR.The experiment is also carried out,and the calculation results show that the residuals of the least square curve fitting is 0.012 μm,which verify the effectiveness of the new algorithm.

We introduce a simple residual pressure self-measurement method for the Fabry-Perot (F-P) cavity of optical MEMS pressure sensor. No extra installation is required and the structure of the sensor is unchanged. In the method, the relationship between residual pressure and external pressure under the same diaphragm deflection condition at different temperatures is analyzed by using the deflection formula of the circular plate with clamped edges and the ideal gas law. Based on this, the residual pressure under the flat condition can be obtained by pressure scanning process and calculation process. We carried out the experiment to compare the residual pressures of two batches MEMS sensors fabricated by two kinds of bonding process. The measurement result indicates that our approach is reliable enough for the measurement.

With the superiority of anti-electromagnetic interference, corrosion resistance, light quality, small size and so on, optical fiber sensing technology is widely used in aerospace industry, petrochemical engineering, power electronics, civil engineering and biological medicine. It can be divided as discrete and distributed. Discrete optical fiber sensing utilizes fiber sensitive element as sensors to detect the quantity to be measured. Optical spectrum, light intensity and polarization are usually used as the sensitivity parameter because they can be modulated by parameter such as rotation, acceleration, electromagnetic field, temperature, pressure, stress, stress, vibration, humidity, viscosity, refractive index and so on. Fiber works as the channel and links the fiber sensor and demodulating equipment. After a long period of research, the discrete optical fiber sensing technology stretch out many branches, we discussed the most representative ones as follows, the fiber grating sensing technique, the fiber fabry perot sensing technique, the fiber gyroscope sensing technique, the fiber intracavity sensing technique, the fiber surface plasma sensing technique, hollow-core fiber whispering gallery mode sensing technique, magnetic fluid fiber sensing technique and fiber-based optical coherence tomography sensing technique. Based on optical effect as rayleigh scattering, Raman scattering and Brillouin scattering, distributed fiber sensing system uses fiber itself as a sensor, when the vibration, stress, voice or temperature acts on the fiber changes, the optical signal transfers inside the fiber will change accordingly. The fiber distributes in a large range and a long distance, then the signal can be located at different positions and realize the multi-position measurement. We discussed the main distributed fiber sensing technologies as follows, the interferometric disturbance fiber sensing technology, the optical frequency domain reflectometry fiber sensing technology, the -optical time domain reflectometer fiber sensing technology, the optical fiber Brillouin sensing technology and the optical fiber Raman sensing technology. The development of technology is promoting the integration and network of optical fiber sensing, now it also becomes a research hotspot. Fiber optic smart sensor network is formed by various discrete and discrete optical fiber sensors in certain topological structure with the function of self-diagnosis and self-healing. Current research concentrates in the following areas, the increase of the multiplex sensor number, the topological structure with higher robustness and the intelligent control of sensing network. In this paper, we discuss the origination, development and research progress of discrete, distributed optical fiber sensing technologies and optical fiber sensing network technology, and the future research direction is also prospected.