@inproceedings{kawasaki-etal-2022-revisiting,
title = "Revisiting Statistical Laws of Semantic Shift in {R}omance Cognates",
author = {Kawasaki, Yoshifumi and
Salingre, Ma{\"e}lys and
Karpinska, Marzena and
Takamura, Hiroya and
Nagata, Ryo},
booktitle = "Proceedings of the 29th International Conference on Computational Linguistics",
month = oct,
year = "2022",
address = "Gyeongju, Republic of Korea",
publisher = "International Committee on Computational Linguistics",
url = "https://aclanthology.org/2022.coling-1.13",
pages = "141--151",
abstract = "This article revisits statistical relationships across Romance cognates between lexical semantic shift and six intra-linguistic variables, such as frequency and polysemy. Cognates are words that are derived from a common etymon, in this case, a Latin ancestor. Despite their shared etymology, some cognate pairs have experienced semantic shift. The degree of semantic shift is quantified using cosine distance between the cognates{'} corresponding word embeddings. In the previous literature, frequency and polysemy have been reported to be correlated with semantic shift; however, the understanding of their effects needs revision because of various methodological defects. In the present study, we perform regression analysis under improved experimental conditions, and demonstrate a genuine negative effect of frequency and positive effect of polysemy on semantic shift. Furthermore, we reveal that morphologically complex etyma are more resistant to semantic shift and that the cognates that have been in use over a longer timespan are prone to greater shift in meaning. These findings add to our understanding of the historical process of semantic change.",
}
1. What is it?
They analyze the cross-lingual semantic shifts of words.
2. What is amazing compared to previous works?
Much research about the lexical semantic change of words has been conducted in monolinguals.
In this paper, the authors try to analyze the semantic shift of word pairs in different languages (French, Italian, and Spanish) derived from the same word (Latin).
3. Where is the key to technologies and techniques?
frequent words (romance, pair of words) are stable: contrary to the raw of innovation (Hamilton+2016) (they cannot assert this with full confidence because they only use documents from recent years)
0. Paper
@inproceedings{kawasaki-etal-2022-revisiting, title = "Revisiting Statistical Laws of Semantic Shift in {R}omance Cognates", author = {Kawasaki, Yoshifumi and Salingre, Ma{\"e}lys and Karpinska, Marzena and Takamura, Hiroya and Nagata, Ryo}, booktitle = "Proceedings of the 29th International Conference on Computational Linguistics", month = oct, year = "2022", address = "Gyeongju, Republic of Korea", publisher = "International Committee on Computational Linguistics", url = "https://aclanthology.org/2022.coling-1.13", pages = "141--151", abstract = "This article revisits statistical relationships across Romance cognates between lexical semantic shift and six intra-linguistic variables, such as frequency and polysemy. Cognates are words that are derived from a common etymon, in this case, a Latin ancestor. Despite their shared etymology, some cognate pairs have experienced semantic shift. The degree of semantic shift is quantified using cosine distance between the cognates{'} corresponding word embeddings. In the previous literature, frequency and polysemy have been reported to be correlated with semantic shift; however, the understanding of their effects needs revision because of various methodological defects. In the present study, we perform regression analysis under improved experimental conditions, and demonstrate a genuine negative effect of frequency and positive effect of polysemy on semantic shift. Furthermore, we reveal that morphologically complex etyma are more resistant to semantic shift and that the cognates that have been in use over a longer timespan are prone to greater shift in meaning. These findings add to our understanding of the historical process of semantic change.", }
1. What is it?
They analyze the cross-lingual semantic shifts of words.
2. What is amazing compared to previous works?
Much research about the lexical semantic change of words has been conducted in monolinguals. In this paper, the authors try to analyze the semantic shift of word pairs in different languages (French, Italian, and Spanish) derived from the same word (Latin).
3. Where is the key to technologies and techniques?
4. How did evaluate it?
They try to reveal the relationship between the degree of semantic change about word pairs in different languages (French, Italian, and Spanish) and:
using linear regression
Table 3 shows that:
5. Is there a discussion?
6. Which paper should read next?