The Spectral Underpinning of word2vec

Jaffe, Ariel; Kluger, Yuval; Lindenbaum, Ofir; Patsenker, Jonathan; Peterfreund, Erez; Steinerberger, Stefan

doi:10.3389/fams.2020.593406

Cited by 5 publications

(3 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, during the COVID-19 pandemic, Twitter has become a place for people to express their opinions, thoughts, and feelings via tweets [5]. A phrase Cite this article as: A. N. Sutranggono and E. M. Imah, "Tweets Emotions Analysis of Community Activities Restriction as COVID-19 Policy in Indonesia Using Support Vector Machine", CommIT Journal 17 (1), [13][14][15][16][17][18][19][20][21][22][23][24][25]2023. conveyed through a tweet can reflect how a person's feelings or emotions are experienced.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Tweets Emotions Analysis of Community Activities Restriction as COVID-19 Policy in Indonesia Using Support Vector Machine

Sutranggono¹,

Imah²

2023

CommIT (Communication and Information Technology) Journal

View full text Add to dashboard Cite

With the rising number of COVID-19 cases in Indonesia, the government has implemented the Imposition of Restrictions on Emergency Community Activities (Pemberlakuan Pembatasan Kegiatan Masyarakat - PPKM) as Indonesia’s COVID-19 policy. Several controversies and protests have colored the implementation of this emergency policy. Some netizens on Twitter voice their opinions about the policy in their tweets. Emotions in tweets can be recognized through text-based emotion detection or emotion analysis. However, textbased emotion detection is a challenging task. One of the main issues in classifying text with a machine learningbased approach deals with the feature dimensions. As a result, appropriate methods for accurately identifying emotion based on the text are required. The research studies an emotions analysis task on Indonesians’ PPKMrelated tweets to understand their emotional state while implementing the PPKM. The machine learning classification algorithms used are Support Vector Machine (SVM) and random forest. The total number of tweets is 4,401. The results show that SVM with linear kernel function combined with the TF-IDF and Chi-Square methods outperforms other classifiers with an accuracy of 0.7528. The accuracy value is higher than those obtained by previous studies. Moreover, the results of the emotion classification on PPKM tweets reveal that most Indonesians are unhappy with the implementation of the PPKM policy.

show abstract

mentioning

confidence: 99%

“…The word embedding method is one of the feature engineering methods commonly used in Natural Language Processing (NLP). Each word in the text can be represented numerically based on text corpora using the word2vec embedding method [16]. Nevertheless, this method has a limitation.…”

mentioning

confidence: 99%

Tweets Emotions Analysis of Community Activities Restriction as COVID-19 Policy in Indonesia Using Support Vector Machine

Sutranggono¹,

Imah²

2023

CommIT (Communication and Information Technology) Journal

View full text Add to dashboard Cite

show abstract

“…For this proof, we derive a relation between latent tree models and a classic result from spectral graph theory known as Fiedler's theorem of nodal domains [18]. This theorem is important in various learning tasks such as clustering data [50], graph partitioning [11], and low dimensional embeddings [25]. To the best of our knowledge, this is the first guarantee derived for spectral partitioning in the setting of latent tree models.…”

Section: Contributions and Outlinementioning

confidence: 99%

Spectral Top-Down Recovery of Latent Tree Models

Aizenbud¹,

Jaffe²,

Wang³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Modeling the distribution of high dimensional data by a latent tree graphical model is a common approach in multiple scientific domains. A common task is to infer the underlying tree structure given only observations of the terminal nodes. Many algorithms for tree recovery are computationally intensive, which limits their applicability to trees of moderate size. For large trees, a common approach, termed divide-and-conquer, is to recover the tree structure in two steps. First, recover the structure separately for multiple randomly selected subsets of the terminal nodes. Second, merge the resulting subtrees to form a full tree. Here, we develop Spectral Top-Down Recovery (STDR), a divide-and-conquer approach for inference of large latent tree models. Unlike previous methods, STDR's partitioning step is non-random. Instead, it is based on the Fiedler vector of a suitable Laplacian matrix related to the observed nodes. We prove that under certain conditions this partitioning is consistent with the tree structure. This, in turn leads to a significantly simpler merging procedure of the small subtrees. We prove that STDR is statistically consistent, and bound the number of samples required to accurately recover the tree with high probability. Using simulated data from several common tree models in phylogenetics, we demonstrate that STDR has a significant advantage in terms of runtime, with improved or similar accuracy.

show abstract

Performance Analysis of Embedding Methods for Deep Learning-Based Turkish Sentiment Analysis Models

Ba Alawi,

Bozkurt

2024

Arab J Sci Eng

View full text Add to dashboard Cite

The complex syntactic structure of Turkish text makes sentiment analysis in natural language processing (NLP) a challenging task. Conventional sentiment analysis methods often fail to effectively identify attitudes in Turkish texts, creating an urgent need for more efficient approaches. To fill this need, our study investigates the effectiveness of embedding techniques including pre-trained Turkish models such as Word2Vec, GloVe, and FastText in addition to two character-level embedding methods, namely, character-integer embedding (CIE) and character one-hot encoding embedding (COE), in conjunction with deep learning models specifically long short-term memory (LSTM), convolution neural networks (CNNs), bidirectional LSTM (Bi-LSTM), and hybrid models, for Turkish short-texts sentiment analysis. DL-based models were investigated on two datasets (e.g., an original Twitter (X) dataset and an accessible hotel reviews dataset). In addition to providing an intensive performance analysis of different embedding strategies and assessing their efficacy in dealing with the linguistic intricacies of Turkish, this study proposed a previously unexplored method in Turkish text representation that relies on a character-level one-hot encoding technique. The obtained findings indicate positive progress using a novel approach utilizing a dual-pathway architecture for both character level and word level that constitutes a substantial contribution to the area of natural language processing (NLP), specifically in the context of complex morphological languages. By employing a hybrid strategy that combines character and word levels on Twitter (X) data, the LSTM model obtained an F1 score of $$0.835 \pm 0.005$$ 0.835 ± 0.005 concerning cross-validation while CNN-BiLSTM attained the highest F1 Score (0.8392) using holdout validation. This strategy consistently produced modest improvements across the second public dataset (hotel reviews dataset) by emerging as the runner-up embedding technique in effectiveness, surpassed only by FastText. Findings provide practical recommendations for practitioners on how to effectively use sentiment analysis to make informed decisions by introducing an extensive performance analysis of the use of embedding techniques and deep learning models for sentiment analysis in Turkish texts, which is crucial in the current age of data analysis.

show abstract

The Spectral Underpinning of word2vec

Cited by 5 publications

References 9 publications

Tweets Emotions Analysis of Community Activities Restriction as COVID-19 Policy in Indonesia Using Support Vector Machine

Tweets Emotions Analysis of Community Activities Restriction as COVID-19 Policy in Indonesia Using Support Vector Machine

Spectral Top-Down Recovery of Latent Tree Models

Performance Analysis of Embedding Methods for Deep Learning-Based Turkish Sentiment Analysis Models

Contact Info

Product

Resources

About