A Co-Embedding Model with Variational Auto-Encoder for Knowledge Graphs

Knowledge graph (KG) embedding has been widely studied to obtain low-dimensional representations for entities and relations. It serves as the basis for downstream tasks, such as KG completion and relation extraction. Traditional KG embedding techniques usually represent entities/relations as vectors or tensors, mapping them in different semantic spaces and ignoring the uncertainties. The affinities between entities and relations are ambiguous when they are not embedded in the same latent spaces. In this paper, we incorporate a co-embedding model for KG embedding, which learns low-dimensional representations of both entities and relations in the same semantic space. To address the issue of neglecting uncertainty for KG components, we propose a variational auto-encoder that represents KG components as Gaussian distributions. In addition, compared with previous methods, our method has the advantages of high quality and interpretability. Our experimental results on several benchmark datasets demonstrate our model’s superiority over the state-of-the-art baselines.

Mobility in Unsupervised Word Embeddings for Knowledge Extraction—The Scholars’ Trajectories across Research Topics

In the knowledge discovery field of the Big Data domain the analysis of geographic positioning and mobility information plays a key role. At the same time, in the Natural Language Processing (NLP) domain pre-trained models such as BERT and word embedding algorithms such as Word2Vec enabled a rich encoding of words that allows mapping textual data into points of an arbitrary multi-dimensional space, in which the notion of proximity reflects an association among terms or topics. The main contribution of this paper is to show how analytical tools, traditionally adopted to deal with geographic data to measure the mobility of an agent in a time interval, can also be effectively applied to extract knowledge in a semantic realm, such as a semantic space of words and topics, looking for latent trajectories that can benefit the properties of neural network latent representations. As a case study, the Scopus database was queried about works of highly cited researchers in recent years. On this basis, we performed a dynamic analysis, for measuring the Radius of Gyration as an index of the mobility of researchers across scientific topics. The semantic space is built from the automatic analysis of the paper abstracts of each author. In particular, we evaluated two different methodologies to build the semantic space and we found that Word2Vec embeddings perform better than the BERT ones for this task. Finally, The scholars’ trajectories show some latent properties of this model, which also represent new scientific contributions of this work. These properties include (i) the correlation between the scientific mobility and the achievement of scientific results, measured through the H-index; (ii) differences in the behavior of researchers working in different countries and subjects; and (iii) some interesting similarities between mobility patterns in this semantic realm and those typically observed in the case of human mobility.

The Image Annotation Refinement in Embedding Feature Space based on Mutual Information

The ever-increasing size of images has made automatic image annotation one of the most important tasks in the fields of machine learning and computer vision. Despite continuous efforts in inventing new annotation algorithms and new models, results of the state-of-the-art image annotation methods are often unsatisfactory. In this paper, to further improve annotation refinement performance, a novel approach based on weighted mutual information to automatically refine the original annotations of images is proposed. Unlike the traditional refinement model using only visual feature, the proposed model use semantic embedding to properly map labels and visual features to a meaningful semantic space. To accurately measure the relevance between the particular image and its original annotations, the proposed model utilize all available information including image-to-image, label-to-label and image-to-label. Experimental results conducted on three typical datasets show not only the validity of the refinement, but also the superiority of the proposed algorithm over existing ones. The improvement largely benefits from our proposed mutual information method and utilizing all available information.

Computational reconstruction of mental representations using human behavior

Revealing the contents of mental representations is a longstanding goal of cognitive science. However, there is currently no general framework for providing direct access to representations of high-level visual concepts. We asked participants to indicate what they perceived in images synthesized from random visual features in a deep neural network. We then inferred a mapping between the semantic features of their responses and the visual features of the images. This allowed us to reconstruct the mental representation of virtually any common visual concept, both those reported and others extrapolated from the same semantic space. We successfully validated 270 of these reconstructions as containing the target concept in a separate group of participants. The visual-semantic mapping uncovered with our method further generalized to new stimuli, participants, and tasks. Finally, it allowed us to reveal how the representations of individual observers differ from each other and from those of neural networks.

Self-Attention Networks and Adaptive Support Vector Machine for Aspect-Level Sentiment Classification

Aspect-level sentiment classification aims to integrating the context to predict the sentiment polarity of aspect-specific in a text, which has been quite useful and popular, e.g. opinion survey and products’ recommending in e-commerce. Many recent studies exploit a Long Short-Term Memory (LSTM) networks to perform aspect-level sentiment classification, but the limitation of long-term dependencies is not solved well, so that the semantic correlations between each two words of the text are ignored. In addition, traditional classification model adopts SoftMax function based on probability statistics as classifier, but ignores the words’ features in the semantic space. Support Vector Machine (SVM) can fully use the information of characteristics and it is appropriate to make classification in the high dimension space, however which just considers the maximum distance between different classes and ignores the similarities between different features of the same classes. To address these defects, we propose the two-stages novel architecture named Self Attention Networks and Adaptive SVM (SAN-ASVM) for aspect-level sentiment classification. In the first-stage, in order to overcome the long-term dependencies, Multi-Heads Self Attention (MHSA) mechanism is applied to extract the semantic relationships between each two words, furthermore 1-hop attention mechanism is designed to pay more attention on some important words related to aspect-specific. In the second-stage, ASVM is designed to substitute the SoftMax function to perform sentiment classification, which can effectively make multi-classifications in high dimensional space. Extensive experiments on SemEval2014, SemEval2016 and Twitter datasets are conducted, compared experiments prove that SAN-ASVM model can obtains better performance.

Query Sense Discovery Approach to Realize the User's Search Intent

The main goal of information retrieval is getting the most relevant documents to a user’s query. So, a search engine must not only understand the meaning of each keyword in the query but also their relative senses in the context of the query. Discovering the query meaning is a comprehensive and evolutionary process; the precise meaning of the query is established as developing the association between concepts. The meaning determination process is modeled by a dynamic system operating in the semantic space of WordNet. To capture the meaning of a user query, the original query is reformulating into candidate queries by combining the concepts and their synonyms. A semantic score characterizing the overall meaning of such queries is calculated, the one with the highest score was used to perform the search. The results confirm that the proposed "Query Sense Discovery" approach provides a significant improvement in several performance measures.

Determination of the content of a construct "volunteer activity" in educational practice: the experience of theoretical comprehension and empirical reconstruction

Introduction. Volunteering is becoming more and more important in the life of civil society and for the formation of human capital, but at the same time there is a lack of comprehensive scientific practice-oriented psychological-pedagogical research on the concept of "volunteer activity". The purpose of this work is to determine the actual effectiveness of an interdisciplinary study of the psychological-pedagogical concept of “volunteer activity” in the educational space. Methods. Empirical data were obtained using the methods of psychosemantics and psycholinguistics (verbal self-descriptions "My volunteer activity" (schoolchildren and students) and "I am an organizer of volunteer activities" (teachers-organizers, leaders of volunteer centers/teams) were used; compilation of an associative dictionary, a "Personal Differential" technique. The data underwent the procedure of content analysis with the extraction of semantic units. To process the results obtained, an iterative method of cluster analysis – k-means clustering – was used; a comparative analysis of fields was implemented. The empirical study involved schoolchildren (N = 44, aged 14-17) and university students (N = 432, aged 18-20) engaged in volunteering and organizers of volunteer activities (N = 37, aged 35-40) from the Altai Territory. The total sample consisted of 910 people. Research results. The scheme of the interdisciplinary study has been tested, which makes it possible to construct the semantic space of a psychological-pedagogical concept of "volunteer activity". For each sample of respondents, two clusters were obtained: schoolchildren and university students (p≤0.01) – the measure of proximity varies from 4.59 to 9.52 for the first cluster and from 4.19 to 7.57 – for the second one; teachers-organizers (p≤0.05) – the measure of proximity varies from 2.67 to 4.89 for the first cluster and from 1.72 to 3.36 – for the second one; volunteer leaders (p≤0.05) – the measure of proximity varies from 2.67 to 4.33 for the first cluster and from 2.35 to 4.73 – for the second one. Significant differences and points of contact of semantic fields were revealed. The average values of the scales according to the "Personal differential" technique are as follows: for schoolchildren – 5, for university students – 16. Conclusion. The prospects for the implementation of this study are outlined, which consist in the empirical generalization of the continuity of implementation of volunteer practices; the structural analysis of psychological-pedagogical technologies and interactive forms of education, the development of a set of tools for assessing the effectiveness of the results of implementation of volunteer programs, the creation of an electronic database of the most effective and adequate practices.

LEXICAL AND PHRASEOLOGICAL PARADIGMATICS OF THE FRENCH LANGUAGE: MADELEINE DE PROUST

The article is devoted to the linguocultural analysis of the idiom madeleine de Proust in modern French. The analysis is based on lexicographic data, component analysis method and text semantic analysis. Using the method of linguocultural decoding they find out emotional and sensual, ethical, aesthetic informations; archetypal, mythological, religious, philosophical and scientific informations, which are the deep foundations of the meaning of a phraseological unit. In the unit madeleine de Proust , the basic types of information are religious and philosophical. The first is associated with the polysemantic inner form of the word Madeleine. It is a nominative name - it means the name of a flour product and a proper name - Madeleine was the name of the girl who invented this culinary dish; this word is associated with the images of Magdalen-sinner and St. Mary-Magdalene, which is reflected in the phraseological system of the French language; with proper name Saint-Jacques, which is used to describe the subject of the material world, and accompanies the writer’s life, the hero in the novel. The philosophical type of information is associated with Proust’s reflections, signs, philosophy, with the concepts of the voluntary and involuntary memory. And all these were pushed by the sensual-emotional information. The polysemous word madeleine in the semantic space of the literary text thanks to the talent of the writer, his emotional and sensual experiences, his stream of consciousness, expanded its meaning and became one of the components of the new language unit madeleine de Proust .

Neurodiversity and (Semantic) Space for the Academic Inclusion of People on the Autism Spectrum

There is a need to rethink functioning and the role of universities that implement inclusive education, understood as high-quality education for everybody, available at all levels of education because of the increasing number of neurodiverse people (with ADHD, autism, dyslexia and other disorders classified as neurodevelopmental).. The aim of our hermeneutical work is an attempt to identify opportunities and limitations on an empirical and theoretical level for creating conditions for the inclusion of students defined as neurodiversity. The research material consists of published own and other authors' studies, and the direction of exploration is determined by the following questions: Are there theoretical and empirical premises for the claim of full inclusion in the academic education of neurodiverse students? What are the research-related limitations that constitute a barrier to the academic inclusion of neurodifferent adolescents? The theoretical background of our work is the theories of social constructivism as defined by Alfred Schűtz, Peter Berger and Thomas Luckmann. The research revealed theoretical and empirical premises confirming the validity of the claim regarding the inclusion of neurodiverse students in academic education due to the intellectual potential of young people, their high self-awareness and the need to provide a growing number of neurodiverse students with conditions for maturing to self-determination in adult life. In Poland, but also in other European countries, the number of students diagnosed with an autism spectrum is not monitored at the national level. Single studies conducted in Poland indicate the similarity of the problems of this group of students with the results of explorations carried out in other countries, and include dropout during the first year of studies, difficulties in relationships with peers, a feeling of loneliness, and a low level of employment after graduation. A large number of people with autism spectrum does not study at all. One of the barriers may be the availability of higher education, which is related to the cultural and economic status. This aspect has not been taken into account in Polish and international research. There was also little dissemination of the idea of neurodiversity, which may be important for the perception of students with the autism spectrum.

Zero-Shot Learning for IMU-Based Activity Recognition Using Video Embeddings

The Activity Recognition Chain generally precludes the challenging scenario of recognizing new activities that were unseen during training, despite this scenario being a practical and common one as users perform diverse activities at test time. A few prior works have adopted zero-shot learning methods for IMU-based activity recognition, which work by relating seen and unseen classes through an auxiliary semantic space. However, these methods usually rely heavily on a hand-crafted attribute space which is costly to define, or a learnt semantic space based on word embedding, which lacks motion-related information crucial for distinguishing IMU features. Instead, we propose a strategy to exploit videos of human activities to construct an informative semantic space. With our approach, knowledge from state-of-the-art video action recognition models is encoded into video embeddings to relate seen and unseen activity classes. Experiments on three public datasets find that our approach outperforms other learnt semantic spaces, with an additional desirable feature of scalability, as recognition performance is seen to scale with the amount of data used. More generally, our results indicate that exploiting information from the video domain for IMU-based tasks is a promising direction, with tangible returns in a zero-shot learning scenario.