Modern logistics companies face significant challenges in efficiently managing dynamic transportation systems, particularly in addressing Dynamic Vehicle Routing Problems (DVRP). These complex problems require specialized expertise for effective algorithm design. Traditional approaches often rely on manual algorithm design for specific routing problems. Reinforcement learning (RL)-based methods, though capable of learning heuristics for diverse routing problems, suffer from poor training efficiency and weak generalization to out-of-distribution (OOD) scenarios. To address these limitations, we leverage the strong generalization capabilities of large language models (LLMs) and adopt an LLM-based heuristic learning approach for DVRP. Our method learns heuristics and autonomously generates executable code within an evolutionary framework, requiring fewer than 10 samples for training. This enables fast training while maintaining robust performance on large-scale OOD instances. Comprehensive experiments across multiple routing problems—including Vehicle Routing Problems (VRP), VRP with Time Windows (VRPTW), DVRP, and DVRP with Time Windows (DVRPTW)—demonstrate that our approach learns effective heuristics and consistently surpasses Greedy baselines. Moreover, in constrained optimization tasks (VRPTW and DVRPTW), our method attains higher feasibility rates than Greedy baselines and approaches the performance of expert-designed heuristics. The proposed method presents a promising and scalable solution with significant potential for real-world industrial deployment.1

Cascade Ranking is a prevalent architecture in large-scale top-k selection systems like recommendation and advertising platforms. Traditional training methods focus on single-stage optimization, neglecting interactions between stages. Recent advances have introduced interaction-aware training paradigms, but still struggle to 1) align training objectives with the goal of the entire cascade ranking (i.e., end-to-end recall of ground-truth items) and 2) learn effective collaboration patterns for different stages. To address these challenges, we propose LCRON, which introduces a novel surrogate loss function derived from the lower bound probability that ground truth items are selected by cascade ranking, ensuring alignment with the overall objective of the system. According to the properties of the derived bound, we further design an auxiliary loss for each stage to drive the reduction of this bound, leading to a more robust and effective top-k selection. LCRON enables end-to-end training of the entire cascade ranking system as a unified network. Experimental results demonstrate that LCRON achieves significant improvement over existing methods on public benchmarks and industrial applications, addressing key limitations in cascade ranking training and significantly enhancing system performance.

Recent advancements in large language models (LLMs) underscore the need for more comprehensive evaluation methods to accurately assess their reasoning capabilities. Existing benchmarks are often domain-specific and thus cannot fully capture an LLM’s general reasoning potential. To address this limitation, we introduce the **Knowledge Orthogonal Reasoning Gymnasium (KORGym)**, a dynamic evaluation platform inspired by KOR-Bench and Gymnasium. KORGym offers over fifty games in either textual or visual formats and supports interactive, multi-turn assessments with reinforcement learning scenarios. Using KORGym, we conduct extensive experiments on 19 LLMs and 8 VLMs, revealing consistent reasoning patterns within model families and demonstrating the superior performance of closed-source models. Further analysis examines the effects of modality, reasoning strategies, reinforcement learning techniques, and response length on model performance. We expect KORGym to become a valuable resource for advancing LLM reasoning research and developing evaluation methodologies suited to complex, interactive environments.

Deep Reinforcement Learning (DRL)-based frameworks, featuring Transformer-style policy networks, have demonstrated their efficacy across various Vehicle Routing Problem (VRP) variants. However, the application of these methods to the Multi-Trip Time-Dependent Vehicle Routing Problem (MTTDVRP) with maximum working hours constraints—a pivotal element of urban logistics—remains largely unexplored. This article introduces a DRL-based method called the Simultaneous Encoder and Dual Decoder Attention Model (SED2AM), tailored for the MTTDVRP with maximum working hours constraints. The proposed method introduces a temporal locality inductive bias to the encoding module of the policy networks, enabling it to effectively account for the time dependency in travel distance/time. The decoding module of SED2AM includes a vehicle selection decoder that selects a vehicle from the fleet, effectively associating trips with vehicles for functional multi-trip routing. Additionally, this decoding module is equipped with a trip construction decoder leveraged for constructing trips for the vehicles. This policy model is equipped with two classes of state representations, fleet state, and routing state, providing the information needed for effective route construction in the presence of maximum working hours constraints. Experimental results using real-world datasets from two major Canadian cities not only show that SED2AM outperforms the current state-of-the-art DRL-based and metaheuristic-based baselines but also demonstrate its generalizability to solve larger scale problems.

Cardiovascular diseases (CVDs) have long been a serious threat to human health and longevity. Cathepsins are involved in a variety of human physiological processes and have a significant impact on the function of the cardiovascular system. Studies have demonstrated an association between cathepsins and cardiovascular disease, and we further investigated whether there are risk and protective factors between various cathepsins and cardiovascular diseases by using a Mendelian randomization approach. This study included 9 cathepsins (B, E, F, G, H, O, S, Z, L2) as exposures, 9 cardiovascular diseases (coronary atherosclerosis, essential hypertension, arrhythmia, angina pectoris, aplastic anemia, heart failure, stroke, malignant lymphoma, and myocardial infarction), and 3 anthropometric indices (Platelet count, diastolic blood pressure, systolic blood pressure) as outcomes. For the Mendelian randomization results, the inverse variance weighting (IVW) method was mainly used, and in order to ensure the accuracy of the results, we carried out the tests of multiplicity and heterogeneity, and the sensitivity analysis using the leave-one-out method. According to the results of MR analysis, cathepsin E is a risk factor for Coronary atherosclerosis (OR=1.0033, 95%CI=1.001-1.0056, P=0.0051) and Myocardial infarction (OR=1.0553, 95%CI=1.0131-1.0993, P=0.0097) and a protective factor for Platelet count (OR=0.9836, 95%CI=0.9728-0.9944, P=0.0031). Cathepsin Z is a risk factor for Essential hypertension (OR=1.0005, 95%CI=1.0000-1.0009, P=0.0439). Cathepsin L2 is a protective factor in chronic heart failure (OR=0.9255, 95%CI=0.8768-0.9769, P=0.005). Cathepsin B is a protective factor for systolic blood pressure (OR=0.801, 95%CI=0.6490-0.9885, P=0.0387).

OBJECTIVE: Data mining of Chinese herbal prescriptions for the treatment of chronic pelvic inflammatory disease of the qi stagnation and blood stasis type yielded the pattern of medication, providing reference for clinical application. METHODS: Literature on the treatment of qi stagnation and blood stasis-type chronic pelvic inflammatory disease by traditional Chinese medicine was collected from China National Knowledge Infrastructure, Wanfang Data, China Biology Medicine disc, and China Science and Technology Journal Database since the establishment of the database and screened for those that met the criteria, to establish a database on the treatment of qi stagnation and blood stasis-type chronic pelvic inflammatory disease by traditional Chinese medicine. The frequency, efficacy, sex and flavour attribution of Chinese medicines were analyzed using Microsoft Office Excel 2019 and Origin 2022 software, and the association rule analysis and cluster analysis of high-frequency medicines were performed using IBM SPSS Modeler 18.0 and IBM SPSS Statistics 27. Results: A total of 112 articles were included in the literature, involving 144 Chinese herbal tastes of traditional Chinese medicines, with a total frequency of use of 1412 times, among which the medicines used ⩾ 35 times were Angelica sinensis, Cyperus rotundus, Fumaric, Red peony, Ligusticum wallichii, Licorice, Peach kernel, Salvia miltiorrhiza, Fructus Aurantii, Safflower, Radix bupleuri, and Peony bark. The drug tastes were mainly warm and bitter, the attributive meridians were mainly liver, spleen, and stomach, and the efficacy was mainly to activate blood circulation and remove blood stasis, regulate qi, tonify the deficiency, and clear heat. The results of association rule analysis showed that the core drug pairs were 'Ligusticum wallichii - Angelica sinensis', 'Licorice - Angelica sinensis', etc. and the commonly used combinations were 'Ligusticum wallichii, Angelica sinensis -Red peony', 'Licorice, Angelica sinensis - Red peony', etc. The HF drugs were clustered into 7 classes by systematic cluster analysis. CONCLUSION: In this paper, the preliminary summary of Chinese medicine treatment of qi stagnation and blood stasis type of chronic pelvic inflammatory disease of the law of medication, treatment of this disease is mostly from the theory of the liver and spleen, to activate blood and eliminate blood stasis, regulate qi to stop pain as the main rule of treatment, the use of medication is mostly warm and cold, to activate blood and eliminate blood stasis medicines are mainly focused on regulating qi to activate blood, dredge the liver to resolve depression.

Objective:Using data mining technology to analyze the pattern of acupuncture points for treating simple obesity of spleen deficiency type to provide reference for clinical treatment of this disease.Methods:The clinical research literature on acupuncture for the treatment of simple obesity of spleen deficiency type was searched by computer from the CNKI, WF, VIP and CBM since the establishment of the database up to March 2024, and the frequency descriptive analysis of the prescription involving acupoints was performed by using Miscrosoft Excel 2021, and the IBM SPSS Modeler 18.0 and SPSS Statistics 27.0 for association rule analysis and cluster analysis of high-frequency acupoints.Results:A total of 126 papers were included, involving 134 acupoints, with a total frequency of 1526 uses. The top five high-frequency acupoints were Zhongwan, Tianshu, Zusanli, Fenglong, and Qihai in order, and the frequency of the above five acupoints was more than 80 times. The meridians were mainly selected from the Stomach Meridian, Ren Meridian and Spleen Meridian. Specific acupoints were mainly used as front Mu acupoint, Back-shu acupoint and Five-shu acupoint. The distribution of acupoints was mainly in the abdomen and lower limbs. Association rule analysis and cluster analysis formed a core point group based on Tianshu-Zhongwan-Fenglong.Conclusion:Acupuncture treatment of simple obesity of spleen deficiency type is based on Zhongwan, Tianshu and Fenglong as the core acupoint combination.

Low Earth orbit (LEO) satellites-based non-terrestrial networks (NTN) are envisioned to complement the fifth-generation (5G) terrestrial networks (TN), enabling global cellular services. However, the high mobility and large coverage of these satellites result in frequent and numerous inter-satellite handovers, leading to signalling storms that degrade the satellite gNodeB services. To address this, we mathematically formulate the handover problem and propose a novel signalling load-aware handover protocol based on conditional handover. We evaluate the effectiveness of the protocol using a customized discrete-event simulator and compare it against a set of baseline conditional handover schemes. Our findings show that the proposed protocol significantly reduces signalling peaks and balances the load more effectively, enhancing the robustness and efficiency of handover in 5G NTN. The simulator is made publicly available.

Dispatching large-scale fleets has been one of the fundamental aspects of managing heterogeneous truckload logistics. This operation involves optimization, visualization, and reporting of the resource plans meticulously crafted by expert planners and dispatchers on a daily basis. However, the limitations of human dispatchers, including errors in communication, routing, compliance, load planning, maintenance oversight, and neglect of driver preferences, can lead to lower customer satisfaction. We present FleetWiz, a Large Language Model-based (LLM) platform that enables logistics industry dispatchers to receive optimal recommendations based on real-time spatial data. FleetWiz seamlessly connects dispatchers, drivers, and resources, centralizing information within a unified resource-request network. This leads to enhanced transit times, reduced delays, and adaptive responses to dynamic conditions. It can execute tasks in different domains including filtering, optimizing, and answering questions based on the network of resources and requests. Specifically, a local Llama3 model is equipped with access to a geo-database for filtering, five different optimization methods for generating plans, and knowledge about the entire network and operations inside the company. Lastly, the tool's reliability, a generic interface for applying LLM agents alongside spatio-temporal optimization models, is demonstrated. The optimization models handle complex dispatching tasks requiring sequential reasoning, allowing the LLM to provide well-informed feedback based on the results.

The heterogeneous fleet and demand vehicle routing problem with time-window constraints (HFDVRPTW) is a crucial optimization problem of significant importance in real-world logistics operations. In this paper, we propose a deep reinforcement learning (DRL)-based method, termed spatial Edge-Feature EnhanCed mulTIgraph fusion encoder With spectral-based embedding and hieRarchical decOder with learnable TEmpoRal positional embedding (EFECTIW-ROTER, pronounced "Effective Router"), to tackle this complex and practical optimization problem. EFECTIW-ROTER utilizes two sparse graphs to represent node connectivity, where nodes correspond to customers and the depot. This sparsity results from the time-window constraints and customers' demand relative to the list of acceptable vehicle attributes specified for service within a heterogeneous fleet, determined by the reachability of the nodes based on these two factors. Leveraging two graph Transformer models, EFECTIW-ROTER's encoding module captures the interactions between the nodes based on these factors. One model encodes customers' heterogeneous demand with spatial edge features based on travel time between the nodes, while the second employs temporal positional embeddings to capture temporal relationships based on time-window ordering. A fusion model is introduced to integrate node interactions based on these graphs. Additionally, a spectral-attention-based pooling ensures effective state representation for the DRL-based method. EFECTIW-ROTER features a hierarchical attention decoder operating in two stages: heterogeneous vehicle selection and node selection. Enhanced with positional embeddings, the decoder is empowered to make effective routing decisions based on time-window constraints' ordering. Experimental results using real-world traffic data from two major Canadian cities confirm EFECTIW-ROTER's better performance over current state-of-the-art DRL-based and heuristic methods. EFECTIW-ROTER reduces travel times while also achieving faster computational times when compared to conventional heuristics. Additional experiments demonstrate its generalizability across larger instances.

Cascade ranking is widely used for large-scale top-k selection problems in online advertising and recommendation systems, and learning-to-rank is an important way to optimize the models in cascade ranking. Previous works on learning-to-rank usually focus on letting the model learn the complete order or top-k order, and adopt the corresponding rank metrics (e.g. OPA and NDCG@k) as optimization targets. However, these targets can not adapt to various cascade ranking scenarios with varying data complexities and model capabilities; and the existing metric-driven methods such as the Lambda framework can only optimize a rough upper bound of limited metrics, potentially resulting in sub-optimal and performance misalignment. To address these issues, we propose a novel perspective on optimizing cascade ranking systems by highlighting the adaptability of optimization targets to data complexities and model capabilities. Concretely, we employ multi-task learning to adaptively combine the optimization of relaxed and full targets, which refers to metrics Recall@m@k and OPA respectively. We also introduce permutation matrix to represent the rank metrics and employ differentiable sorting techniques to relax hard permutation matrix with controllable approximate error bound. This enables us to optimize both the relaxed and full targets directly and more appropriately. We named this method as Adaptive Neural Ranking Framework (abbreviated as ARF). Furthermore, we give a specific practice under ARF. We use the NeuralSort to obtain the relaxed permutation matrix and draw on the variant of the uncertainty weight method in multi-task learning to optimize the proposed losses jointly. Experiments on a total of 4 public and industrial benchmarks show the effectiveness and generalization of our method, and online experiment shows that our method has significant application value.

OBJECTIVE: To analyze the regularity of traditional Chinese medicine compound in the treatment of primary hypotension by data mining technique, and to study the mechanism of core drugs on primary hypotension by network pharmacology and molecular docking. METHODS: The databases of China National Knowledge Infrastructure (CNKI), Wanfang Data Knowledge Service Platform and Ancient and Modern Medical Records Cloud Platform were searched, the prescriptions for the treatment of primary hypotension were selected, and the prescriptions for the treatment of primary hypotension were established using Excel 2021. The relevant prescriptions selected in the literature were subjected to frequency statistics and association rule analysis to obtain the core drug pairs. The components and targets of the core drug pair were searched in the TCMSP database, intersected with the disease targets of primary hypotension in the Geneccards and PharmgKb databases, and a visual network was constructed by Cytoscape 3.10.0 software, and molecular docking validation was performed using AutoDockTools software, and finally the molecular docking results were visualized by Pymol software. Results: 73 prescriptions were obtained, 116 herbs were obtained, the cumulative frequency was 704, and the top ten herbs were Radix Astragali, Radix Angelicae Sinensis, Radix Glycyrrhizae, Radix Codonopsis, Rhizoma Atractylodis Macrocephalae, Cimicifuga racemosa, Radix Bupleuri, Pericarpium Citri Reticulatae, Radix Ophiopogon is After association rule analysis, the core drug pair was finally obtained: Angelica sinensis – Astragalus membranaceus. After screening, 21 core components and 143 targets for the treatment of primary hypotension were obtained, and the core targets were JUN, EGFR, TP53, and MAPK14.Molecular docking showed that quercetin, kaempferol and formononetin had good binding activity to the core target for molecular docking. CONCLUSION: The core components of Angelica sinensis and Astragalus membranaceus in the treatment of primary hypotension can act on the cardiovascular system and regulate blood pressure by regulating blood flow and regulating vasoconstriction and relaxation.

Objective: To investigate the pathogenesis and pathological mechanism of salivation in TCM, and to provide theoretical basis for TCM formula.METHODS: Critical targets associated with salivation were selected by Genecard database and protein-protein interaction (PPI) network analysis, and gene ontology (GO) function and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were performed.The TCMSP database was used to search the TCM components related to the key targets, and the relevant candidate compounds were selected according to oral bioavailability (ob)≥ 30%, drug-like drug (dl) ≥ 0.18, degree, and the molecular binding energy was calculated between the top four targets in degree and each candidate compound pairwise to verify the reliability; Through TCMSP to find the candidate compounds related to traditional Chinese medicine, the medicinal properties, meridians, classification frequency analysis of traditional Chinese medicine, and select the degree value >5 of traditional Chinese medicine as a target-compound-traditional Chinese medicine network. Results: A total of 575 genes were obtained. The top 15 degrees were selected as key targets, and GO functional analysis showed that they were associated with 132 biological processes, 17 cellular components, and 9 molecular functions; Matching to 16 key targets resulted in 34 candidate compounds, molecular docking was performed with the key target of degree first 4 for a total of 136 pairs of molecular docking, and the results showed that the binding activity between the target and the compound was strong and the prediction accuracy was high; The results of frequency statistics showed that the medicinal properties of traditional Chinese medicines that could act on the disease target were mostly cold, the medicinal taste was bitter and sweet, and most of them belonged to the liver, lung meridian, and kidney meridian, and the efficacy types were mainly clearing away heat and tonifying deficiency; the traditional Chinese medicines ranked > 5 degrees were fenugreek, kohua, Patrinia, safflower, Sophora flavescens, Portulaca oleracea, Sophora japonica root, white fruit, and wood butterfly, indicating that these nine traditional Chinese medicines may have a strong therapeutic effect on salivation. Conclusion: This experiment introduces data support at the modern molecular experiment level for TCM pathogenesis theory and prescription for the salivation based on bioinformatics and medication rule, which provides a new breakthrough point for research on modern TCM pathogenesis and prescription theories.

Efficient goods movement is a vital aspect of logistics and urban planning, impacting the flow of goods and the quality of life for residents. To aid in this, we present an interactive map-based system for visualizing and analyzing goods’ movements using GPS traces. The system takes raw GPS signal data and road network data as input, then performs preprocessing and spatial analysis on the data using Flask framework and Python scripts. The system offers a user-friendly interface to explore the patterns of goods movement dynamically. It displays the temporal and spatial movement trips in the city, providing an intuitive way to analyze and optimize goods movement. Our system’s ability to explore and visualize goods movement patterns makes it an essential addition to the existing literature on urban transportation analysis and a valuable tool for logistics companies and urban planners.

Truck speed prediction plays a key role in truck transportation management. However, it is a very challenging task since the truck traffic usually shows complex patterns. Most of the existing traffic prediction methods lack the ability to model the dynamic spatial-temporal correlations of truck traffic or ignore contributing contextual factors that impact traffic. Also, truck traffic data is typically sparse and noisy, which makes the truck speed prediction an even more challenging task. How to improve the truck speed prediction by taking advantage of other relevant truck traffic information (such as the truck flow) has not been investigated in depth. Additionally, traffic congestions and poor driving conditions caused by extreme weather conditions can make sudden changes in the general pattern of the truck speed. In this paper, we propose a novel Multi-Task Context Based Gated Recurrent Unit Graph Convolutional Network (MT-C2G) to predict the truck speed under extreme weather conditions. MT-C2G includes four major components: The spatial dependence learning component captures the spatial dependencies shaped by the topological structure of the road network. Truck traffic feature temporal dependence modeling component is built to acquire the temporal dependencies involved in the truck traffic features, and contextual feature temporal dependence modeling component employs a layer of GRU units to capture the temporal dependencies of contextual factors. The multi-task learning component then leverages the information between the truck speed and flow prediction tasks through attention mechanism for improving the performance. Moreover, a data augmentation method SMOTE is utilized to balance the data with the extreme weather conditions. Experiments on two real datasets demonstrate that the proposed MT-C2G fairly outperforms six state-of-the-art traffic prediction methods.

Most machine learning classifiers only concern classification accuracy, while certain applications (such as medical diagnosis, meteorological forecasting, and computation advertising) require the model to predict the true probability, known as a calibrated estimate. In previous work, researchers have developed several calibration methods to post-process the outputs of a predictor to obtain calibrated values, such as binning and scaling methods. Compared with scaling, binning methods are shown to have distribution-free theoretical guarantees, which motivates us to prefer binning methods for calibration. However, we notice that existing binning methods have several drawbacks: (a) the binning scheme only considers the original prediction values, thus limiting the calibration performance; and (b) the binning approach is non-individual, mapping multiple samples in a bin to the same value, and thus is not suitable for order-sensitive applications. In this paper, we propose a feature-aware binning framework, called Multiple Boosting Calibration Trees (MBCT), along with a multi-view calibration loss to tackle the above issues. Our MBCT optimizes the binning scheme by the tree structures of features, and adopts a linear function in a tree node to achieve individual calibration. Our MBCT is non-monotonic, and has the potential to improve order accuracy, due to its learnable binning scheme and the individual calibration. We conduct comprehensive experiments on three datasets in different fields. Results show that our method outperforms all competing models in terms of both calibration error and order accuracy. We also conduct simulation experiments, justifying that the proposed multi-view calibration loss is a better metric in modeling calibration error. In addition, our approach is deployed in a real-world online advertising platform; an A/B test over two weeks further demonstrates the effectiveness and great business value of our approach.

Low-resource stylized sequence-to-sequence (S2S) generation is in high demand. However, its development is hindered by the datasets which have limitations on scale and automatic evaluation methods. We construct two large-scale, multiple-reference datasets for low-resource stylized S2S, the Machine Translation Formality Corpus (MTFC) that is easy to evaluate and the Twitter Conversation Formality Corpus (TCFC) that tackles an important problem in chatbots. These datasets contain context to source style parallel data, source style to target parallel data, and non-parallel sentences in the target style to enable the semi-supervised learning. We provide three baselines, the pivot-based method, the teacher-student method, and the back-translation method. We find that the pivot-based method is the worst, and the other two methods achieve the best score on different metrics.

Conventional approaches for formality style transfer borrow models from neural machine translation, which typically requires massive parallel data for training. However, the dataset for formality style transfer is considerably smaller than translation corpora. Moreover, we observe that informal and formal sentences closely resemble each other, which is different from the translation task where two languages have different vocabularies and grammars. In this paper, we present a new approach, Sequence-to-Sequence with Shared Latent Space (S2S-SLS), for formality style transfer, where we propose two auxiliary losses and adopt joint training of bi-directional transfer and auto-encoding. Experimental results show that S2S-SLS (with either RNN or Transformer architectures) consistently outperforms baselines in various settings, especially when we have limited data.

Text summarization plays an important role in various NLP applications. Using templates with generation methods is an effective way to address abstractive summarization. However, existing template-enhanced generation approaches use templates in a naive way and mainly adopt RNN-based Seq2Seq models, so they cannot make full use of valid information in the templates and suffer from templates' noise. To mitigate these problems, we propose a new abstractive summarization model called Summarization Transformer with Template-aware Representation (STTR), which uses a template-aware document encoding module and a document representation shifting loss to preserve the useful information and filter the noise of the template. The experiments on the Gigaword and LCSTS datasets show that our method outperforms baseline models and achieves a new state-of-the-art.

Open domain response generation has achieved remarkable progress in recent years, but sometimes yields short and uninformative responses. We propose a new paradigm, prototypethen-edit for response generation, that first retrieves a prototype response from a pre-defined index and then edits the prototype response according to the differences between the prototype context and current context. Our motivation is that the retrieved prototype provides a good start-point for generation because it is grammatical and informative, and the post-editing process further improves the relevance and coherence of the prototype. In practice, we design a contextaware editing model that is built upon an encoder-decoder framework augmented with an editing vector. We first generate an edit vector by considering lexical differences between a prototype context and current context. After that, the edit vector and the prototype response representation are fed to a decoder to generate a new response. Experiment results on a large scale dataset demonstrate that our new paradigm significantly increases the relevance, diversity and originality of generation results, compared to traditional generative models. Furthermore, our model outperforms retrieval-based methods in terms of relevance and originality.

Yunli Wang, Yu Wu, Lili Mou, Zhoujun Li, Wenhan Chao. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.

Detecting changes within an unfolding event in real time from news articles or social media enables to react promptly to serious issues in public safety, public health or natural disasters. In this study, we use on-line Latent Dirichlet Allocation (LDA) to model shifts in topics, and apply on-line change point detection (CPD) algorithms to detect when significant changes happen. We describe an on-line Bayesian change point detection algorithm that we use to detect topic changes from on-line LDA output. Extensive experiments on social media data and news articles show the benefits of on-line LDA versus standard LDA, and of on-line change point detection compared to off-line algorithms. This yields F-scores up to 52% on the detection of significant real-life changes from these document streams.

The goal of keyphrase extraction is to automatically identify the most salient phrases from documents. The technique has a wide range of applications such as rendering a quick glimpse of a document, or extracting key content for further use. While previous work often assumes keyphrases are a static property of a given documents, in many applications, the appropriate set of keyphrases that should be extracted depends on the set of documents that are being considered together. In particular, good keyphrases should not only accurately describe the content of a document, but also reveal what discriminates it from the other documents. In this paper, we study this problem of extracting discriminative keyphrases. In particularly, we propose to use the hierarchical semantic structure between candidate keyphrases to promote keyphrases that have the right level of specificity to clearly distinguish the target document from others. We show that such knowledge can be used to construct better discriminative keyphrase extraction systems that do not assume a static, fixed set of keyphrases for a document. We show how this helps identify key expertise of authors from their papers, as well as competencies covered by online courses within different domains.

BACKGROUND: Expression quantitative trait loci (eQTL) mapping is often used to identify genetic loci and candidate genes correlated with traits. Although usually a group of genes affect complex traits, genes in most eQTL mapping methods are considered as independent. Recently, some eQTL mapping methods have accounted for correlated genes, used biological prior knowledge and applied these in model species such as yeast or mouse. However, biological prior knowledge might be very limited for most species. RESULTS: We proposed a data-driven prior knowledge guided eQTL mapping for identifying candidate genes. At first, quantitative trait loci (QTL) analysis was used to identify single nucleotide polymorphisms (SNP) markers that are associated with traits. Then co-expressed gene modules were generated and gene modules significantly associated with traits were selected. Prior knowledge from QTL mapping was used for eQTL mapping on the selected modules. We tested and compared prior knowledge guided eQTL mapping to the eQTL mapping with no prior knowledge in a simulation study and two barley stem rust resistance case studies. The results in simulation study and real barley case studies show that models using prior knowledge outperform models without prior knowledge. In the first case study, three gene modules were selected and one of the gene modules was enriched with defense response Gene Ontology (GO) terms. Also, one probe in the gene module is mapped to Rpg1, previously identified as resistance gene to stem rust. In the second case study, four gene modules are identified, one gene module is significantly enriched with defense response to fungus and bacterium. CONCLUSIONS: Prior knowledge guided eQTL mapping is an effective method for identifying candidate genes. The case studies in stem rust show that this approach is robust, and outperforms methods with no prior knowledge in identifying candidate genes.

Semi-supervised clustering incorporating biological relevance as a prior knowledge has been favored over the past decade. However, selection of prior knowledge has been a challenge. We generate prior knowledge from Gene Ontology (GO) terms at different levels of GO hierarchy and use them to study their impact on the performance of subsequent clustering of microarray data by using MPCKMeans and GOFuzzy. We evaluate the performance by F-measure and the number of specific GO terms and transcription factors. The clustering result with prior knowledge generated from lower levels of GO hierarchy have higher F-measure and more number of specific GO terms and transcription factors. MPCKMeans with prior knowledge generated from multiple levels in the GO hierarchy outperforms GOFuzzy with prior knowledge from the first level in the GO hierarchy. A small amount (1-2%) of prior knowledge can improve semi-supervised clustering result substantially and the more specific prior knowledge is generally more efficient in guiding the semi-supervised clustering process.

More and more public data sets which contain information about individuals are published in recent years. The urgency to reduce the risk of privacy disclosure from such data sets makes the approaches of privacy protection for data publishing widely employed. There are two popular models for privacy protection: k-anonymity and l-diversity. k-anonymity focuses on reducing the probability of identifying a particular person, which requires that each equivalence class (a set of records with the same identifier attributes) contains at least k records. l-diversity concentrates on reducing the inference from released sensitive attributes. It requires that each equivalence class has at least l “well-represented” sensitive attribute values. In this study, we view the privacy protection problem in a brand new perspective. We proposed a new model, Unique Distinct l-SR diversity based on the sensitivity of private information. Also, we presented two performance measures to evaluate how much sensitive information can be inferred from an equivalence class. l-SR diversity algorithm was implemented to achieve Unique Distinct l-SR diversity. We tested l-SR diversity on one benchmark data set and two synthetic data sets, and compared it with other l-diversity algorithms. The results show that our algorithm achieved better performance on minimizing inference of sensitive information and reached the comparable generalization data quality compared with other data publishing algorithms.

For omni-directional imaging based cooperative robot vision system, panorama unrolling is an important problem. We present an eight direction symmetry reuse algorithm for this problem, principles of this algorithm are: 1) Treat the original image to be unrolled as a series of co-centric circles, and uniformly partition them into eight parts of symmetrical sector regions. 2) According to symmetry principle in spatial geometry, we construct the symmetrical transform relations pixel coordinates among these regions. So, we need compute only one region of pixel coordinate for panorama unrolling, by use of complex ray-trace coordinate mapping, while the other seven parts of regions can be determined by symmetrically reusing results from the first region, which reduces seven-eighths computational burden for ray-tracing method. Experiments indicate that, under the same precision, our algorithm improves 3.28 times of unrolling speed compared with ray-trace unrolling method. In omnidirectional video with look-up table method, the idea of eight direction symmetry reuse can also be used to reduce the look-up table size to only one-eighth of the original table.

This paper proposes a new texture description metric, the color distribution evenness (CDE) measure, and discusses its usage of performing multi-scale texture analysis. Further, CDE measure is applied to color-texture segmentation of natural images, upon which we propose the ISBEC (Image Segmentation Based on the distribution Evenness of Colors) algorithm. Experiments verify the effectiveness of CDE measure for texture analysis and that of ISBEC for color-texture segmentation.

Internet is used as one of major sources of health information. However, some studies show that the readability of health information presented on health web sites is difficult for many consumers. Readability formulas usually measure difficulty of writing style, instead of difficulty of content. In order to recommend health information with appropriate reading level to consumers, we investigate the feasibility of identifying text difficulty of health information using machine learning methods. Support Vector Machine is used to classify consumer health information into easy to read and reading level for the general public. Three feature sets: surface linguistic features, word difficulty features, unigrams and their combinations are compared in terms of classification accuracy. Unigram features alone reach an accuracy of 80.71%, and the combination of three feature sets is the most effective in classification with accuracy of 84.06%. They are significantly better than surface linguistic features, word difficulty features and their combination.

A rough mill production system is an unpredictable dynamic system for which the effects on production of many factors are strongly interrelated. A distributed decision support system for dynamic jag (a load of lumber) selection in defect sensitive production is proposed. Case-Based Reasoning and heuristic rules determine recommended jags for cut-lists in the decision process. Using a statistical method, cases are grouped by categorizing problems and solutions. Heuristic rules measure the similarity between current and past cases. The framework for a decision support system has two layers in which jag types are selected at the top level and specific jags are chosen on the bottom level. Decisions from different sources influence both top level and bottom level. The dynamic character of the system is taken into account by collaboration between distributed decision points that can change indices by retrieving previous cases and storing new cases. The recommended jags are tested by simulation and used for feedback to the decision support system. Finally, the system is validated by comparison with actual production.