Semantic models for assessing the influence of a combination of factors on the development of diseases

The analysis of approaches and solutions to the problem of risk assessment and prognosis of conditions and development of diseases is presented. It is shown that the implementation of software services on various platforms complicates the possibility of their comprehensive use and the choice between the available solutions. This has risen the urgency of creating a unified semantic model of diseases that integrates various methods and approaches to solving this problem and accumulates knowledge about risks and prognosis in a unified information space. A new semantic model is proposed to take into account influence of a combination of factors on development of various events that threaten health and life. The feature of the model is its independence from a specific disease or a group of diseases, which allows it to be used in various branches of medicine. This model has been tested on the IACPaaS platform. A software solver has been implemented that allows generating a clear explanation based on the knowledge base and analysis of the patient's electronic medical record. The application of the new model for the formation of knowledge is shown on the example of risk assessment and prognosis of cardiovascular events.

Building upon the Foundations of Branching Space-Times

This chapter introduces a variety of events that are definable in BST and discusses in which histories these events occur. This gives rise to the concept of the occurrence proposition for events of various kinds. Of particular interest are transitions, defined as pairs of events, one of which is appropriately below the other. Transitions play a crucial role in later chapters. The chapter then discusses the topological aspects of BST, which are picked up again in Chapter 9. It defines a natural topology for BST: the diamond topology, and describes some important facts about it, focusing on the Hausdorff property and local Euclidicity. The chapter also gives an overview of how BST structures may be used to build semantic models for languages with temporal and modal operators.

A Word Selection Method for Producing Interpretable Distributional Semantic Word Vectors

Distributional semantic models represent the meaning of words as vectors. We introduce a selection method to learn a vector space that each of its dimensions is a natural word. The selection method starts from the most frequent words and selects a subset, which has the best performance. The method produces a vector space that each of its dimensions is a word. This is the main advantage of the method compared to fusion methods such as NMF, and neural embedding models. We apply the method to the ukWaC corpus and train a vector space of N=1500 basis words. We report tests results on word similarity tasks for MEN, RG-65, SimLex-999, and WordSim353 gold datasets. Also, results show that reducing the number of basis vectors from 5000 to 1500 reduces accuracy by about 1.5-2%. So, we achieve good interpretability without a large penalty. Interpretability evaluation results indicate that the word vectors obtained by the proposed method using N=1500 are more interpretable than word embedding models, and the baseline method. We report the top 15 words of 1500 selected basis words in this paper.

The Austrian Semantic EO Data Cube Infrastructure

Big optical Earth observation (EO) data analytics usually start from numerical, sub-symbolic reflectance values that lack inherent semantic information (meaning) and require interpretation. However, interpretation is an ill-posed problem that is difficult for many users to solve. Our semantic EO data cube architecture aims to implement computer vision in EO data cubes as an explainable artificial intelligence approach. Automatic semantic enrichment provides semi-symbolic spectral categories for all observations as an initial interpretation of color information. Users graphically create knowledge-based semantic models in a convergence-of-evidence approach, where color information is modelled a-priori as one property of semantic concepts, such as land cover entities. This differs from other approaches that do not use a-priori knowledge and assume a direct 1:1 relationship between reflectance values and land cover. The semantic models are explainable, transferable, reusable, and users can share them in a knowledgebase. We provide insights into our web-based architecture, called Sen2Cube.at, including semantic enrichment, data models, knowledge engineering, semantic querying, and the graphical user interface. Our implemented prototype uses all Sentinel-2 MSI images covering Austria; however, the approach is transferable to other geographical regions and sensors. We demonstrate that explainable, knowledge-based big EO data analysis is possible via graphical semantic querying in EO data cubes.

Asindetske rečenice kao jezičko-stilska dominanta pripovijedaka Jovana Radulovića

The paper pinpoints and describes asyndetic sentences as the linguistic and stylistic dominant of Jovan Radulović’s short stories. The analysis was primarily syntac- tic-semantic, because its goal was to single out and describe the basic structural-semantic models of asyndetic sentences in Radulović’s literary work. The method of analysis was analytical-synthetic. The analysis of asyndetic sentences in Jovan Radulović’s short stories greatly chang- es the view on the syntactic-semantic and stylistic status of these sentences in the language in general, and in the literary-artistic style in particular. Namely, Jovan Radulović shows originality and innovation by creating as many as five structural-semantic types of asyndetic sentences. Thus Radulović forms asyndetic sentences: 1) whose clauses combine narrative and direct speech as the speech of literary characters, 2) whose clauses combine narrative and free indirect speech, 3) whose clauses represent sentences of different functional goal or purpose, 4) which combine clauses expressed by predicate and non-verbal statements, and 5) whose asyndetic clauses allow“insertion” into the structure of another asyndetic clause. And it is exactly these types that represent the main argument that asyndetic sentences are structured according to the principles of the (bound) text, and not according to the princi- ples of a complex sentence. The analysis also showed that Jovan Radulović often includes a syndetic clause in the structure of polyclause asyndetic sentences in the mesophoric or epiphoric position for semantic and / or stylistic reasons, thus forming an asyndetic-syndetic sentence, which represents a comparative basis for declaring asyndetic sentences as a stylistic device. The analysis also showed that Jovan Radulović uses three orthographic signs for syntactic delimitation of clauses without conjuntions in the asyndetic sentence, namely commas, dashes and semicolons, where the comma is the most common and structurally- stylistically unmarked sign, while dash and semicolon are always used intentionally: for a special or structural or semantic, or stylistic emphasis on the role of one of the clauses within the whole asyndetic sentence.

Introducing Various Semantic Models for Amharic: Experimentation and Evaluation with Multiple Tasks and Datasets

The availability of different pre-trained semantic models has enabled the quick development of machine learning components for downstream applications. However, even if texts are abundant for low-resource languages, there are very few semantic models publicly available. Most of the publicly available pre-trained models are usually built as a multilingual version of semantic models that will not fit well with the need for low-resource languages. We introduce different semantic models for Amharic, a morphologically complex Ethio-Semitic language. After we investigate the publicly available pre-trained semantic models, we fine-tune two pre-trained models and train seven new different models. The models include Word2Vec embeddings, distributional thesaurus (DT), BERT-like contextual embeddings, and DT embeddings obtained via network embedding algorithms. Moreover, we employ these models for different NLP tasks and study their impact. We find that newly-trained models perform better than pre-trained multilingual models. Furthermore, models based on contextual embeddings from FLAIR and RoBERTa perform better than word2Vec models for the NER and POS tagging tasks. DT-based network embeddings are suitable for the sentiment classification task. We publicly release all the semantic models, machine learning components, and several benchmark datasets such as NER, POS tagging, sentiment classification, as well as Amharic versions of WordSim353 and SimLex999.

Deep neural networks and visuo-semantic models explain complementary components of human ventral-stream representational dynamics

Deep neural networks (DNNs) are promising models of the cortical computations supporting human object recognition. However, despite their ability to explain a significant portion of variance in neural data, the agreement between models and brain representational dynamics is far from perfect. Here, we address this issue by asking which representational features are currently unaccounted for in neural timeseries data, estimated for multiple areas of the human ventral stream via source-reconstructed magnetoencephalography (MEG) data. In particular, we focus on the ability of visuo-semantic models, consisting of human-generated labels of higher-level object features and categories, to explain variance beyond the explanatory power of DNNs alone. We report a gradual transition in the importance of visuo-semantic features from early to higher-level areas along the ventral stream. While early visual areas are better explained by DNN features, higher-level cortical dynamics are best accounted for by visuo-semantic models. These results suggest that current DNNs fail to fully capture the visuo-semantic features represented in higher-level human visual cortex and suggest a path towards more accurate models of ventral stream computations.