The increasing availability of low-cost wearable devices and smartphones has significantly advanced the field sensor-based human activity recognition (HAR), attracting considerable research interest. One major challenges in HAR is domain shift problem cross-dataset recognition, which occurs due to variations users, device types, sensor placements between source dataset target dataset. Although adaptation methods have shown promise, they typically require access during training process, might...

Accurate road networks play a crucial role in modern mobile applications such as navigation and last-mile delivery. Most existing studies primarily focus on generating open areas like main roads avenues, but little attention has been given to the generation of community closed residential areas, which becomes more significant due growing demand for door-to-door services food This lack research is attributed challenges related sensing data availability quality. In this paper, we design novel...

In recent years, emergency last-mile logistics (ELML) have played an essential role in urban emergencies. The efficient allocation of couriers ELML is practical significance to ensure the supply materials, especially public health emergencies (PHEs). However, becomes challenging due instability demand, dynamic comprehension, and evolutional delivery environment for caused by PHEs. While existing work has delved into allocation, impact PHEs on demand-supply-delivery yet be considered. this...

The popularity of online e-commerce has promoted the rapid development last-mile logistics in recent years. In services, to ensure delivery efficiency and enhance user experience, zone is proposed perform task assignment, which a fundamental part delivery. Each courier responsible for one zone. Couriers will collect orders belonging their zones from station deliver customers. Existing partition practices consist manual experience-based static optimization-based methods, order amount...

To reduce the difficulty and enhance enthusiasm of private-owned electric vehicles (EVs) to participate in frequency regulation ancillary service market (FRASM), a decision aid model (DAM) is proposed.This paper presents three options for EV participating FRASM, i. e., base mode (BM), unidirectional charging (UCM), bidirectional charging/discharging (BCDM), based on reasonable simplification users' willingness.In BM, individual EVs will not be involved DAM assist users set optimal schemes...

Logistics audience expansion, the process for logistics companies to find potential long-term customers, is one of most important tasks business growth. However, existing methods conventional expansion fall short due two significant challenges, intricate interplay multiple complex factors in scenario and emphasis on service usage instead one-time promotions. To address above limitations, we design LOGAE-TKG, a method based temporal knowledge graph, which consists three components: (i) graph...

Points of Interest (POIs), such as entertainment, dining, and living, are crucial for urban planning location-based services. However, the high dynamics expensive updating costs POIs pose a key roadblock their applications. This is especially true developing countries, where active economic activities lead to frequent POI updates (e.g., merchants closing down new ones opening). Therefore, updating, i.e., detecting different names same (alias) update database, has become an urgent but...

Abstract Clothing-change person re-identification is an emerging research topic aimed at reidentifying individuals who have changed their clothing. This highly challenging and has not received sufficient attention to date. Most existing methods primarily focus on clothing features, but in real-world scenarios where change attire, the identification accuracy of conventional networks significantly decreases. The key challenge how effectively extract clothing-agnostic human features. paper...

Prediction of couriers' delivery timely rates in advance is essential to the logistics industry, enabling companies take preemptive measures ensure normal operation services. This becomes even more critical during anomaly conditions like epidemic outbreak, which rate will decline markedly and fluctuates significantly. Existing studies pay less attention scenario. Moreover, many works focusing on prediction tasks scenarios fail explicitly model abnormal events, e.g., treating external factors...

Fine-grained community-building information, such as building names and accurate geographical coordinates, is critical for a range of practical applications like navigation door-to-door services (e.g., on-demand delivery last-mile delivery). A common practice traditional methods to gather information usually relies on manual collection, which typically labor-intensive time-consuming. To address these issues, we utilize the massive data generated from e-commerce design framework, AutoBuild,...

Retrieval-augmented generation integrates the capabilities of large language models with relevant information retrieved from an extensive corpus, yet encounters challenges when confronted real-world noisy data. One recent solution is to train a filter module find content but only achieve suboptimal noise compression. In this paper, we propose introduce bottleneck theory into retrieval-augmented generation. Our approach involves filtration by simultaneously maximizing mutual between...

Large language models (LLMs) have demonstrated strong reasoning capabilities. Nevertheless, they still suffer from factual errors when tackling knowledge-intensive tasks. Retrieval-augmented represents a promising approach. However, significant challenges persist, including inaccurate and insufficient retrieval for complex questions, as well difficulty in integrating multi-source knowledge. To address this, we propose Beam Aggregation Reasoning, BeamAggR, framework multi-hop QA. BeamAggR...

Accurate prediction of order transportation time is essential for customer satisfaction in logistics. Existing methods based on origin-destination (OD) pairs do not consider the diversity road segments, while route-based may fail to account real-time traffic conditions due infrequent dispatch schedules logistics vehicles. In reality, e-commerce platforms have collaborated with multiple companies parcel delivery, providing a richer dataset that offers more comprehensive view conditions. The...

Recovering the fine-grained working process of couriers is becoming one essential problems for improving express delivery systems because knowing detailed how accomplish their daily work facilitates analyzing, understanding, and optimizing procedure. Although coarse-grained courier trajectories waybill time data can be collected, this problem still challenging due to noisy with spatio-temporal biases, lacking ground truth couriers’ behaviors, complex correlations between behaviors. Existing...