In reality, multi-view multi-label learning has a richer descriptive power than single-view single-label learning. However, due to multiple uncertainties in practical situations, the process of collecting data often results in missing views and labels. Therefore, the double incomplete multi-view multi-label classification task emerged. In this complex field, two issues can be characterized as follows: i) How to embed complementary learning objectives in the view feature space ? ii) How to jointly learn the consistency and complementarity of views? To this end, this paper proposes a View-Channel Mixer Network (VCMN) based on the mixing operations of a multilayer perceptron to address this complex problem. First, we designed a View-Channel Mixer module to model the complementary relationships in the high-dimensional feature space of the views. Second, cross-view and cross-sample contrastive loss is used to enhance the network’s ability to learn consistent representations across multiple views, enabling the joint learning of complementarity and consistency. Our method demonstrates competitive experimental results compared to other state-of-the-art methods on five widely used multi-view multi-label datasets.

Abstract Unlike Western languages, word segmentation is necessary for Japanese sentences because they do not have word boundaries. The performances of existing morphological analyzers for Japanese sentences are very high. However, it is difficult to segment sentences mostly written in Hiragana, which is a Japanese writing system simpler than Kanji, because clues to segment the sentences decrease. In this study, we created a word segmentation model of Hiragana sentences using two types of BERT: unigram and bigram BERT models. We pre-trained the BERT models with Wikipedia and fine-tuned them with the core data of the Balanced Corpus of Contemporary Written Japanese for word segmentation. In addition to the two types of BERT-based word segmentation systems, we developed a word segmentation system for Hiragana sentences using KyTea, a toolkit developed for analyzing text, with a focus on languages requiring word segmentation. We compared them in word segmentation of Hiragana sentences. The experiments revealed that the unigram BERT-based word segmentation system outperformed the bigram BERT-based word segmentation system and the KyTea-based word segmentation system.

The authors compared two methods for annotating a corpus for the named entity (NE) recognition task using non-expert annotators: (i) revising the results of an existing NE recognizer and (ii) manually annotating the NEs completely. The annotation time, degree of agreement, and performance were evaluated based on the gold standard. Because there were two annotators for one text for each method, two performances were evaluated: the average performance of both annotators and the performance when at least one annotator is correct. The experiments reveal that semi-automatic annotation is faster, achieves better agreement, and performs better on average. However, they also indicate that sometimes, fully manual annotation should be used for some texts whose document types are substantially different from the training data document types. In addition, the machine learning experiments using semi-automatic and fully manually annotated corpora as training data indicate that the F-measures could be better for some texts when manual instead of semi-automatic annotation was used. Finally, experiments using the annotated corpora for training as additional corpora show that (i) the NE recognition performance does not always correspond to the performance of the NE tag annotation and (ii) the system trained with the manually annotated corpus outperforms the system trained with the semi-automatically annotated corpus with respect to newswires, even though the existing NE recognizer was mainly trained with newswires.

We propose a new spam detection technique using the text clustering based on vector space model. Our method computes disjoint clusters automatically using a spherical k-means algorithm for all spam/non-spam mails and obtains centroid vectors of the clusters for extracting the cluster description. For each centroid vectors, the label ('spam' or 'non-spam') is assigned by calculating the number of spam email in the cluster. When new mail arrives, the cosine similarity between the new mail vector and centroid vector is calculated. Finally, the label of the most relevant cluster is assigned to the new mail. By using our method, we can extract many kinds of topics in spam/non-spam email and detect the spam email efficiently. In this paper, we describe the our spam detection system and show the result of our experiments using the Ling-Spam test collection.