Contrastive Trajectory Learning for Tour Recommendation

The main objective of Personalized Tour Recommendation (PTR) is to generate a sequence of point-of-interest (POIs) for a particular tourist, according to the user-specific constraints such as duration time, start and end points, the number of attractions planned to visit, and so on. Previous PTR solutions are based on either heuristics for solving the orienteering problem to maximize a global reward with a specified budget or approaches attempting to learn user visiting preferences and transition patterns with the stochastic process or recurrent neural networks. However, existing learning methodologies rely on historical trips to train the model and use the next visited POI as the supervised signal, which may not fully capture the coherence of preferences and thus recommend similar trips to different users, primarily due to the data sparsity problem and long-tailed distribution of POI popularity. This work presents a novel tour recommendation model by distilling knowledge and supervision signals from the trips in a self-supervised manner. We propose Contrastive Trajectory Learning for Tour Recommendation (CTLTR), which utilizes the intrinsic POI dependencies and traveling intent to discover extra knowledge and augments the sparse data via pre-training auxiliary self-supervised objectives. CTLTR provides a principled way to characterize the inherent data correlations while tackling the implicit feedback and weak supervision problems by learning robust representations applicable for tour planning. We introduce a hierarchical recurrent encoder-decoder to identify tourists’ intentions and use the contrastive loss to discover subsequence semantics and their sequential patterns through maximizing the mutual information. Additionally, we observe that a data augmentation step as the preliminary of contrastive learning can solve the overfitting issue resulting from data sparsity. We conduct extensive experiments on a range of real-world datasets and demonstrate that our model can significantly improve the recommendation performance over the state-of-the-art baselines in terms of both recommendation accuracy and visiting orders.

Accelerating DNN Training Through Selective Localized Learning

Training Deep Neural Networks (DNNs) places immense compute requirements on the underlying hardware platforms, expending large amounts of time and energy. We propose LoCal+SGD, a new algorithmic approach to accelerate DNN training by selectively combining localized or Hebbian learning within a Stochastic Gradient Descent (SGD) based training framework. Back-propagation is a computationally expensive process that requires 2 Generalized Matrix Multiply (GEMM) operations to compute the error and weight gradients for each layer. We alleviate this by selectively updating some layers' weights using localized learning rules that require only 1 GEMM operation per layer. Further, since localized weight updates are performed during the forward pass itself, the layer activations for such layers do not need to be stored until the backward pass, resulting in a reduced memory footprint. Localized updates can substantially boost training speed, but need to be used judiciously in order to preserve accuracy and convergence. We address this challenge through a Learning Mode Selection Algorithm, which gradually selects and moves layers to localized learning as training progresses. Specifically, for each epoch, the algorithm identifies a Localized→SGD transition layer that delineates the network into two regions. Layers before the transition layer use localized updates, while the transition layer and later layers use gradient-based updates. We propose both static and dynamic approaches to the design of the learning mode selection algorithm. The static algorithm utilizes a pre-defined scheduler function to identify the position of the transition layer, while the dynamic algorithm analyzes the dynamics of the weight updates made to the transition layer to determine how the boundary between SGD and localized updates is shifted in future epochs. We also propose a low-cost weak supervision mechanism that controls the learning rate of localized updates based on the overall training loss. We applied LoCal+SGD to 8 image recognition CNNs (including ResNet50 and MobileNetV2) across 3 datasets (Cifar10, Cifar100, and ImageNet). Our measurements on an Nvidia GTX 1080Ti GPU demonstrate upto 1.5× improvement in end-to-end training time with ~0.5% loss in Top-1 classification accuracy.

Multiple weak supervision for short text classification

For short text classification, insufficient labeled data, data sparsity, and imbalanced classification have become three major challenges. For this, we proposed multiple weak supervision, which can label unlabeled data automatically. Different from prior work, the proposed method can generate probabilistic labels through conditional independent model. What’s more, experiments were conducted to verify the effectiveness of multiple weak supervision. According to experimental results on public dadasets, real datasets and synthetic datasets, unlabeled imbalanced short text classification problem can be solved effectively by multiple weak supervision. Notably, without reducing precision, recall, and F1-score can be improved by adding distant supervision clustering, which can be used to meet different application needs.

Kenakalan remaja klithih yang mengarah pada konflik sosial dan kekerasan di Yogyakarta

Tulisan ini bertujuan untuk mendapatkan informasi mengenai peristiwa klithih yang akhir-akhir ini terjadi di Kota Yogyakarta dan sekitarnya. Klithih merupakan jenis kenakalan remaja yang mengarah pada konflik sosial dan kekerasan di masyarakat. Klithih pada awalnya adalah sebuah ajang yang digunakan oleh para remaja untuk menunjukkan eksistensinya di dalam pergaulan antar remaja di Yogyakarta. Pada akhirnya klithih akhirnya berubah menjadi ajang untuk menciptakan sebuah konflik sosial dan kekerasan dengan menyasar siapa saja yang berada di jalan raya. Penyebab umum terjadinya klithih selain untuk menunjukkan eksistensi kelompok remajanya/ peer group juga karena lemahnya pengawasan dan control sosial oleh keluarga dan sekolah karena sebagian besar pelakunya adalah remaja anak sekolah. Dalam perspektif sosiologi, tidak ada jawaban tunggal dalam menjelaskan realitas sosial termasuk fenomena klithih ini karena sosiologi merupakan ilmu sosial berparadigma ganda. Demikian juga dalam menjelaskan realitas klithih di Yogyakarta. Tinjauan klithih di jalanan Kota Yogyakarta ini vital dilakukan agar supaya penjelasan tidak parsial sehingga dapat mengungkapkan pemahaman yang universal dan menyeluruh. Pilihan teoretik tersebut memiliki implikasi metodologis yang selanjutnya diharapkan berakhir pada ditemukannya langkah penyelesaian yang tepat oleh seluruh pihak yang terkait. Langkah solutif untuk pencegahan dan mengatasi terjadinya klithih perlu dilakukan untuk mengembangkan relasi sosial menjadi lebih harmonis dan humanis sekaligus mengurangi terjadinya penyakit sosial yang berupa klithih. This paper aims to get information about klithih events that recently occurred in the city of Yogyakarta and surrounding areas. Klithih is a type of juvenile delinquency that leads to social conflict and violence in society. Klithih was originally an event used by teenagers to show their existence in the association between teenagers in Yogyakarta. Eventually klithih finally turned into a place to create a social conflict and violence by targeting anyone who was on the highway. The most common cause of klithih in addition to showing the existence of adolescents/peer groups is also due to the weak supervision and social control by families and schools because most of the perpetrators are teenage school children. In the perspective of sociology, there is no single answer in explaining social reality including this klithih phenomenon because sociology is a social paradigm with multiple paradigms. Likewise in explaining the reality of klithih in Yogyakarta. This klithih review on the streets of Yogyakarta is vital so that the explanation is not partial so that it can reveal a universal and comprehensive understanding. The theoretical choice has methodological implications which are then expected to end in the discovery of an appropriate settlement step by all parties concerned. Solutive steps to prevent and overcome the occurrence of klithih needs to be done to develop social relations to be more harmonious and humanist while reducing the occurrence of social diseases in the form of klithih.

Preventing The Illegal Trade of Smuggling Small Arms And Light Weapons (SALW) through UNPoA in Indonesia

<em>Action is needed to combat the illegal trade of Small Arms and Light Weapons (SALW) because transnational crime is not easy to commit. However, internal conflicts make Indonesia more vulnerable to firearms smuggling, considering its geographical conditions and weak supervision at the border. Consequently, the government cooperates with neighboring countries to maintain national integrity and safety. Also, the government is active in the international regime to deal with illegal trade of SALW through the United Nations Program of Action. This study is legal research by incorporating primary, secondary, and tertiary data. The results showed that Indonesia's position in the United Nations Program of Action helps prevent firearms smuggling and increase capacity-building assistance.Furthermore, the government collaborates with the Ministry of Foreign Affairs and amends and revises Law Number 8 of 1948 concerning Registration and Granting of Permits for the use of Firearms to prevent illegal trade of SALW. In the regional scope, similar collaboration is also conducted with Southeast countries. This is supported by implementing the PoA to Combat Transnational Crime by holding the ASEAN Ministerial Meeting on Transnational Crime (AMMTC). In the international scope, the United Nations Convention Against Transnational Crime and its three protocols were introduced to eradicate the illegal trade of SAWL.</em>

Weakly Supervised Classification of Hyperspectral Image Based on Complementary Learning

In recent years, supervised learning-based methods have achieved excellent performance for hyperspectral image (HSI) classification. However, the collection of training samples with labels is not only costly but also time-consuming. This fact usually causes the existence of weak supervision, including incorrect supervision where mislabeled samples exist and incomplete supervision where unlabeled samples exist. Focusing on the inaccurate supervision and incomplete supervision, the weakly supervised classification of HSI is investigated in this paper. For inaccurate supervision, complementary learning (CL) is firstly introduced for HSI classification. Then, a new method, which is based on selective CL and convolutional neural network (SeCL-CNN), is proposed for classification with noisy labels. For incomplete supervision, a data augmentation-based method, which combines mixup and Pseudo-Label (Mix-PL) is proposed. And then, a classification method, which combines Mix-PL and CL (Mix-PL-CL), is designed aiming at better semi-supervised classification capacity of HSI. The proposed weakly supervised methods are evaluated on three widely-used hyperspectral datasets (i.e., Indian Pines, Houston, and Salinas datasets). The obtained results reveal that the proposed methods provide competitive results compared to the state-of-the-art methods. For inaccurate supervision, the proposed SeCL-CNN has outperformed the state-of-the-art method (i.e., SSDP-CNN) by 0.92%, 1.84%, and 1.75% in terms of OA on the three datasets, when the noise ratio is 30%. And for incomplete supervision, the proposed Mix-PL-CL has outperformed the state-of-the-art method (i.e., AROC-DP) by 1.03%, 0.70%, and 0.82% in terms of OA on the three datasets, with 25 training samples per class.

Weakly Supervised Faster-RCNN+FPN to classify small animals in camera trap images

<div>Camera traps have revolutionized animal research of many species that were previously nearly impossible to observe due to their habitat or behavior.</div><div>Deep learning has the potential to overcome the workload to the class automatically those images according to taxon or empty images. However, a standard deep neural network classifier fails because animals often represent a small portion of the high-definition images. Therefore, we propose a workflow named Weakly Object Detection Faster-RCNN+FPN which suits this challenge. The model is weakly supervised because it requires only the animal taxon label per image but doesn't require any manual bounding box annotations. First, it automatically performs the weakly supervised bounding box annotation using the motion from multiple frames. Then, it trains a Faster-RCNN+FPN model using this weak supervision.<br></div><div>Experimental results have been obtained on two datasets and an easily reproducible testbed.</div>