This paper explores Multi-Domain Sequential Recommendation (MDSR), an advancement of Multi-Domain Recommendation that incorporates sequential context. Recent MDSR approach exploits domain-specific sequences, decoupled from mixed-domain histories, to model domain-specific sequential preference, and use mixeddomain histories to model domain-shared sequential preference. However, the approach faces challenges in accurately obtaining domain-specific sequential preferences in the target domain, especially when users only occasionally engage with it. In such cases, the history of users in the target domain is limited or not recent, leading the sequential recommender system to capture inaccurate domain-specific sequential preferences. To address this limitation, this paper introduces Multi-Domain Sequential Recommendation via Domain Space Learning (MDSR-DSL). Our approach utilizes cross-domain items to supplement missing sequential context in domain-specific sequences. It involves creating a "domain space" to maintain and utilize the unique characteristics of each domain and a domain-to-domain adaptation mechanism to transform item representations across domain spaces. To validate the effectiveness of MDSR-DSL, this paper extensively compares it with state-of-the-art MD(S)R methods and provides detailed analyses.

Recently, web platforms are operating various service domains simultaneously. Targeting a platform that operates multiple service domains, we introduce a new task, Multi-Domain Recommendation to Attract Users (MDRAU), which recommends items from multiple "unseen" domains with which each user has not interacted yet, by using knowledge from the user's "seen" domains. In this paper, we point out two challenges of MDRAU task. First, there are numerous possible combinations of mappings from seen to unseen domains because users have usually interacted with a different subset of service domains. Second, a user might have different preference for each of the target unseen domains, which requires recommendations to reflect users' preference on domains as well as items. To tackle these challenges, we propose DRIP framework that models users' preference at two levels (i.e., domain and item) and learns various seen-unseen domain mappings in a unified way with masked domain modeling. Our extensive experiments demonstrate the effectiveness of DRIP in MDRAU task and its ability to capture users' domain-level preferences.

Recent studies on one-class classification have achieved a remarkable performance by employing the self-supervised classifier that predicts the type of pre-defined geometric transformations applied on in-class images. However, they cannot correctly identify in-class images as in-class at all when the input images have various viewpoints (e.g., translated or rotated images), because their classification-based in-class scores assume that in-class images always have a fixed viewpoint. Pointing out that humans can easily recognize such images having various viewpoints as the same class, in this work, we aim to propose a one-class classifier robust to geometrically-transformed inputs, named as GROC. To this end, we remark that in-class images match better with the in-class transformations than out-of-class images do. We introduce a conformity score indicating how strongly an input image agrees with one of the predefined in-class transformations, then utilize the conformity score with our proposed agreement measures for one-class classification. Our extensive experiments demonstrate that GROC is able to accurately distinguish in-class images from out-of-class images regardless of whether the inputs are geometrically-transformed or not, whereas the existing methods fail.

Most recent studies on detecting and localizing temporal anomalies have mainly employed deep neural networks to learn the normal patterns of temporal data in an unsupervised manner. Unlike them, the goal of our work is to fully utilize instance-level (or weak) anomaly labels, which only indicate whether any anomalous events occurred or not in each instance of temporal data. In this paper, we present WETAS, a novel framework that effectively identifies anomalous temporal segments (i.e., consecutive time points) in an input instance. WETAS learns discriminative features from the instance-level labels so that it infers the sequential order of normal and anomalous segments within each instance, which can be used as a rough segmentation mask. Based on the dynamic time warping (DTW) alignment between the input instance and its segmentation mask, WETAS obtains the result of temporal segmentation, and simultaneously, it further enhances itself by using the mask as additional supervision. Our experiments show that WETAS considerably outperforms other baselines in terms of the localization of temporal anomalies, and also it provides more informative results than point-level detection methods.

Recent studies on one-class classification have achieved a remarkable performance, by employing the self-supervised classifier that predicts the geometric transformation applied to in-class images. However, they cannot identify in-class images at all when the input images are geometrically-transformed (e.g., rotated images), because their classification-based in-class scores assume that input images always have a fixed viewpoint, as similar to the images used for training. Pointing out that humans can easily recognize such transformed images as the same class, in this work, we aim to propose a one-class classifier robust to geometrically-transformed inputs, named as GROC. To this end, we introduce a conformity score which indicates how strongly an input image agrees with one of the predefined in-class transformations, then utilize the conformity score with our proposed agreement measures for one-class classification. Our extensive experiments demonstrate that GROC is able to accurately distinguish in-class images from out-of-class images regardless of whether the inputs are geometrically-transformed or not, whereas the existing methods fail.

The goal of one-class collaborative filtering (OCCF) is to identify the user-item pairs that are positively-related but have not been interacted yet, where only a small portion of positive user-item interactions (e.g., users' implicit feedback) are observed. For discriminative modeling between positive and negative interactions, most previous work relied on negative sampling to some extent, which refers to considering unobserved user-item pairs as negative, as actual negative ones are unknown. However, the negative sampling scheme has critical limitations because it may choose "positive but unobserved" pairs as negative. This paper proposes a novel OCCF framework, named as BUIR, which does not require negative sampling. To make the representations of positively-related users and items similar to each other while avoiding a collapsed solution, BUIR adopts two distinct encoder networks that learn from each other; the first encoder is trained to predict the output of the second encoder as its target, while the second encoder provides the consistent targets by slowly approximating the first encoder. In addition, BUIR effectively alleviates the data sparsity issue of OCCF, by applying stochastic data augmentation to encoder inputs. Based on the neighborhood information of users and items, BUIR randomly generates the augmented views of each positive interaction each time it encodes, then further trains the model by this self-supervision. Our extensive experiments demonstrate that BUIR consistently and significantly outperforms all baseline methods by a large margin especially for much sparse datasets in which any assumptions about negative interactions are less valid.

Anomaly detection task, which identifies abnormal patterns in data, has been widely applied to various domains. Most recent work on anomaly detection have focused on an accurate modeling of the normal data based on unsupervised methods. To get a satisfactory anomaly detection accuracy, they need pure normal data without abnormal data. This scenario requires many labels to get pure normal data. In many real-world scenarios, there exist abundant unlabeled data and a limited number of partially labeled anomalies. This paper proposes a novel anomaly detection method, PUMAD, which uses a Positive and Unlabeled (PU) learning approach to learn from abundant unlabeled data and a small number of partially labeled anomalies (i.e., positives). PUMAD successfully works on the anomaly detection scenario by exploiting deep metric learning with a hashing-based filtering method. Extensive experimental results on real-world benchmark datasets demonstrate that our approach based on PU learning is effective to detect anomalies. PUMAD achieves a much higher accuracy of up to 24% than state-of-the-art competitors.

Users' behaviors observed in many web-based applications are usually heterogeneous, so modeling their behaviors considering the interplay among multiple types of actions is important. However, recent collaborative filtering (CF) methods based on a metric learning approach cannot learn multiple types of user actions, because they are developed for only a single type of user actions. This paper proposes a novel metric learning method, called METAS, to jointly model heterogeneous user behaviors. Specifically, it learns two distinct spaces: 1) action space which captures the relations among all observed and unobserved actions, and 2) entity space which captures high-level similarities among users and among items. Each action vector in the action space is computed using a non-linear function and its corresponding entity vectors in the entity space. In addition, METAS adopts an efficient triplet mining algorithm to effectively speed up the convergence of metric learning. Experimental results show that METAS outperforms the state-of-the-art methods in predicting users' heterogeneous actions, and its entity space represents the user-user and item-item similarities more clearly than the space trained by the other methods.

Most neural network models for sentiment classification use word vectors pre-trained by word embedding methods to represent a word. Although word vectors are trained on large corpus, most of them are restricted by the vocabularies in the corpus. Since sentiment classification models have to capture subtle meaning of sentence, it is desirable to represent words that have not been pre-trained by word embedding method. To achieve this goal, we propose a sentiment classification model with convolutional neural network using multiple word representations. We represent a word by three embedding methods including word2vec, GloVe, and our method which is based on a character level embedding method that successfully captures subtle differences between words. Experimental results from three datasets show that our model with an additional character level embedding method improves the accuracy of the sentiment classification.

With the popularization of social networking services, numerous words are newly emerging every day in personalized document sources. Slang terms, abbreviations, newly coined words, and non-grammatical words or expressions belong here, and people are more likely to use these words with a certain sentimental tendency compared to other standard words. Thus, it becomes important to find their meanings or sentiments to analyze the sentiment of user-generated texts. This paper proposes a novel sentiment analysis model, termed DualSentiNet, which predicts the sentiments of newly emerged words and documents at the same time. Our model is composed of three parts: (i) a word-level sentiment regression network, (ii) a document-level sentiment classification network, and (iii) a shared word embedding layer. DualSentiNet makes a word embedding layer shared by two different networks, thereby learning richer information about both word-level and document-level sentiments through two-way back-propagation. Consequently, it improves the performance of sentiment prediction by preventing word vectors from being overfitted. Experimental results show that DualSentiNet significantly outperforms competitors in terms of both document sentiment classification accuracy and the word sentiment regression RMSE. In addition, DualSentiNet produces better word embedding by reflecting both word and document sentiments.