In this article we compare different vector models (tf-idf, word2vec, fasttext, lda, lsi, artm) in the short text clustering task, using a dataset of job vacancy descriptions in Russian. A two-step experiment is proposed to determine the best model and its hyperparameters based on the quality of the resulting short text clusters. In the first stage, we investigate how various hyperparameters of each model can affect the clusters, produced by training a K-means model on each of the vector representations. In particular, we consider in detail, how the size of the output vector representation in each of our models can influence the quality of the final clusters. We also provide an extensive analysis of the effects of various regularization options for clusters, learned using the vectors produced by the ARTM algorithm. During the second stage, the models showing the best results in the previous step (word2vec, fasttext) are analyzed in greater detail. We compare the effectiveness of these models against datasets of different sizes, as well as using different structures of the source fragments (partial elements or full texts of vacancy descriptions). In our experiments, the highest quality of clusters (evaluated using the ARI metric) was achieved by word2vec, closely followed by the fasttext model. Finally, we perform a topic analysis for each of the resulting clusters and evaluate their homogeneity.

No abstract

As numbers of educational programmes and courses grow, the need for a method of comparison becomes apparent. In this paper we discuss the overall state of education data mining, the variety of document types and formats used for educational content and propose the combined similarity measure for educational course programmes, Our proposed similarity measure uses three most important in our opinion elements of course programmes-course descriptions, educational results of the course and the structure of the educational course. We describe our approach to calculate similarity for each of this component pairs as well as provide primary experimental results and their evaluation using mean average precision metric.

No abstract

This paper proposes an approach, improving the quality of the original educational course programmes semantic search algorithm, based on vector representations, produced by distributional semantic. The proposed approach works by providing an expert with interpretable topic filtering of courses in search results. Application of probabilistic topic modeling based on additive regularization ensures the interpretability of vector components in representations of texts, allowing the expert, in the process of exploratory search, to narrow down the set of relevant documents found previously by using the vector model. In our experiments, we study the applied task of educational course search, using current requirements of the labor market (requirements described in professional standards serve as search queries). The implementation of topic filtering is based on the open-source library BigARTM. We investigate the influence of hyperparameters and the choice of regularizers in the construction of a topic model on the improvement of quality of educational course semantic search using various vector models: word2vec, fasttext, TF-IDF are investigated.