Testing word embeddings for Polish

Abstract

Distributional Semantics postulates the representation of word meaning in the form of numeric vectors which represent words which occur in context in large text data. This paper addresses the problem of constructing such models for the Polish language. The paper compares the effectiveness of models based on lemmas and forms created with Continuous Bag of Words (CBOW) and skipgram approaches based on different Polish corpora. For the purposes of this comparison, the results of two typical tasks solved with the help of distributional semantics, i.e. synonymy and analogy recognition, are compared. The results show that it is not possible to identify one universal approach to vector creation applicable to various tasks. The most important feature is the quality and size of the data, but different strategy choices can also lead to significantly different results.

Building Semantic User Profile for Polish Web News Portal

“…Considering also the dynamic nature of news content and changing user interests, we verify this claim and use a Word2Vec model trained on a corpora consisting of Polish Wikipedia and National Corpus of Polish Language NKJP [25] to compare with our models built on a much smaller custom corpus with the same model parameters. As the text preprocessing tools used by [3] are different from ours, we compare models trained on all word forms.…”

“…The goal is to verify if using a pre-trained Word2Vec representation on a large external corpus results in a better performance than a much smaller custom text collection. We compare the results for the model published by [25] trained on a large corpus of the Polish language with a custom representation trained on our corpus with the same model parameters.…”

“…However, most of these resources are designed for English: there are few public tools available for the Polish language. A 100D vector of Polish words trained with the skip-gram Word2Vec model has been published by [18], while [25] built a model trained for the Polish language from various resources -Polish Wikipedia and National Polish Language Corpora NKJP [31]. The authors evaluated models built with different parameters and vector dimensions on the tasks of analogy and synonym retrieval and concluded that the choice of the model depends on the task.…”

“…We compared our results with a model trained by [25] built on Polish Wikipedia and NKJP [31] corpora. The model is a skip-gram Word2Vec architecture with hierarchical softmax with a vector size of 100.…”

Improving Time-Series Demand Modeling in Hospitality Business by Analytics of Public Event Datasets

“…Such space can have the sense of words ''meanings'' or ''styles'', depending on details of the process of embeddings production. Following a common practice, we use off the shelf embbedings made from a large national corpus of literary and official documents [38]. Consequently, each sample Y gets translated into matrix Z 30×100 of embeddings of the first thirty title words.…”