Improving semantic change analysis by combining word embeddings and word frequencies

Englhardt, Adrian; Willkomm, Jens; Schäler, Martin; Böhm, Klemens

doi:10.1007/s00799-019-00271-6

Cited by 5 publications

(6 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Yet, there remains a question: did word frequency change affect the change of word meaning, or more specifically, did they present a negative relationship as found by Englhardt et al (2020) This seemingly strange situation may attribute to the setting of this research. Firstly, the general negative relations between word frequency and meaning change found before (Englhardt et al 2020;Hamilton et al 2016) were all concluded from historical corpora, in which the strangeness caused by sudden events like pandemics could be diluted. Secondly, when we take close look at these three hot words, it is clear that while the word 新冠 COVID-19 is a newly coined word due to the pandemic, the other two words are both existing words attached with new meanings and concerns in this pandemic.…”

Section: Correlation Testsmentioning

confidence: 75%

“…Secondly, inspired by co-occurrence rates measuring the correlation strength between any two highlighted words (Zhu et al 2020), we also conducted correlations between frequency ranks of hot words in pairs. And thirdly, informed by Englhardt et al (2020), we inspected correlations between word meaning change ratio (1) and monthly mean word frequency ranks. As defined in (1), the word meaning change ratio is calculated based on the number of new nearest neighbors of a certain word.…”

Section: Research Proceduresmentioning

confidence: 99%

“…For instance, the rate of semantic change negatively correlates with word-usage frequency (Englhardt et al 2019;Hamilton et al 2016). Research applying word embeddings has obtained bountiful findings, however, further innovations are needed in the choice of materials (e.g., Grag et al 2018;Mou et al 2015), which were mostly formal texts from historical corpora (Englhardt et al 2020) and less dynamic compared with instant data from online social media. By using word embeddings and retrieving real linguistic data relevant to certain influential social events on social media, more changes in connotative meaning, as well as traces of associated discourse may be found.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Dynamics of language in social emergency: investigating COVID-19 hot words on Weibo

Zhou¹,

Li²,

Chen³

et al. 2022

Glottometrics

View full text Add to dashboard Cite

Drawing on word embeddings techniques and tracking the frequency and semantic change of hot words on Sina Weibo during the COVID-19 pandemic, this study investigates how language and discourse change during crisis. More specifically, correlation tests were conducted between word frequency ranks, pandemic data, and word meaning change ratio. Results indicated that the frequency of some hot words changed with both pandemic data and the frequency of other hot words, which were significantly correlated with the American pandemic data rather than that of China. Moreover, February of 2020 saw the most distinctive semantic changes marked by a large part of the nearest neighbors for WAR metaphors. The correlations between changes in the frequency and nearest neighbors of COVID-19 related hot words exhibited some acceptable peculiarities. This study proves the availability of studying discourse through language change by observing minor semantic change on connotation level from social media, which adds a new perspective to the impact of the COVID-19 pandemic.

show abstract

Section: Correlation Testsmentioning

confidence: 75%

Section: Research Proceduresmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Dynamics of language in social emergency: investigating COVID-19 hot words on Weibo

Zhou¹,

Li²,

Chen³

et al. 2022

Glottometrics

View full text Add to dashboard Cite

show abstract

“…Finally, in order to further analyze the characteristics and differences in users’ sensory evaluations, this study conducted word frequency analysis on the parks with emotional values greater than 1 based on the results of relative emotional values. The word frequency analysis technology was used to segment, check, and fit the evaluation of the park [ 75 , 76 ], and people’s emotional perceptions were obtained from the five aspects of vision, smell, hearing, touch, and taste, so as to analyze the results of the crowd experience of different waterfront spaces.…”

Section: Methodsmentioning

confidence: 99%

Landscape Aesthetic Value of Waterfront Green Space Based on Space–Psychology–Behavior Dimension: A Case Study along Qiantang River (Hangzhou Section)

Xiao-jia

Chen²,

Huang

et al. 2023

IJERPH

View full text Add to dashboard Cite

As an important part of urban green infrastructure, the landscape effect of the urban waterfront green space varies, and sometimes, the green space with an excellent landscape aesthetic value fails to serve the needs of most citizens. This seriously affects the construction of a green ecological civilization and the implementation of the concept of “common prosperity” in China. Based on multi-source data, this study took the Qiantang River Basin as an example, selected 12 representative waterfront green spaces along the river as the research objects, and used qualitative and quantitative analysis methods to determine the landscape aesthetic value of the research area from the different dimensions of space, psychology, and physiology. We examined the relationship between each dimension so as to objectively and comprehensively reflect the landscape value characteristics of the waterfront green space in the study area and provide a reasonable theoretical framework and practical development path for future urban waterfront green space landscape design. We obtained the following results: (1) The results of the spatial dimension research indicated that the spatial value index of the waterfront green space in the study area was three-dimensional space > vertical space > horizontal space, and the overall spatial value was low; Qianjiang Ecological Park obtained the highest value (0.5473), and Urban Balcony Park obtained the lowest value (0.4619). (2) The results of the psychological dimension indicated that people’s perceptions of the waterfront green space in the study area were relatively weak, mainly focusing on visual perception, but the waterfront green space with a relative emotional value greater than one accounted for 75%, and the overall recognition of the landscape was high. (3) The results of the behavioral dimension showed that the overall heat of the waterfront green space in the study area was insufficient (1.3719–7.1583), which was mainly concentrated in low-heat levels, and the population density was unevenly distributed (0.0014–0.0663), which was mainly concentrated in the medium-density level. The main purpose of users was to visit, and they stayed an average of 1.5 h. (4) The results of the coupling coordination analysis of the spatial–psychological–behavioral dimensions showed that the landscape value of the waterfront green space in the study area presented a form of ‘high coupling degree and low coordination degree’.

show abstract

“…Through this approach, tokens with high intra-year instability will be penalized and vice-verse for more stable tokens. Along with token distance calculations, it has been shown that including token frequency improves results compared to using distance alone [37]. We calculated token frequency as the ratio of token frequency in the more recent year over the frequency of the previous year.…”

Section: Methodsmentioning

confidence: 99%

Changing word meanings in biomedical literature reveal pandemics and new technologies

Nicholson

Alquaddoomi

Rubinetti

et al. 2022

Preprint

View full text Add to dashboard Cite

While we often think of words as having a fixed meaning that we use to describe a changing world, words are also dynamic and changing. Scientific research can also be remarkably fast-moving, with new concepts or approaches rapidly gaining mind share. We examined scientific writing, both preprint and pre-publication peer-reviewed text, to identify terms that have changed and examine their use. One particular challenge that we faced was that the shift from closed to open access publishing meant that the size of available corpora changed by over an order of magnitude in the last two decades. We developed an approach to evaluate semantic shift by accounting for both intra- and inter-year variability using multiple integrated models. Using this strategy and examining year-by-year changes revealed thousands of change points in both corpora. We found change points for tokens including 'cas9', 'pandemic', and 'sars' among many others. The consistent change-points between pre-publication peer-reviewed and preprinted text were largely related to the COVID-19 pandemic. We developed a web app for exploration (https://greenelab.github.io/word-lapse/) that enables users to investigate individual terms. To our knowledge, this analysis is the first to examine the semantic shift in biomedical preprints and pre-publication peer-reviewed text, and it lays the foundation for future work to examine how terms acquire new meaning and the extent to which that process is encouraged or discouraged by peer review.

show abstract

Improving semantic change analysis by combining word embeddings and word frequencies

Cited by 5 publications

References 28 publications

Dynamics of language in social emergency: investigating COVID-19 hot words on Weibo

Dynamics of language in social emergency: investigating COVID-19 hot words on Weibo

Landscape Aesthetic Value of Waterfront Green Space Based on Space–Psychology–Behavior Dimension: A Case Study along Qiantang River (Hangzhou Section)

Changing word meanings in biomedical literature reveal pandemics and new technologies

Contact Info

Product

Resources

About