Improved NER on Turkish tweets using semi-supervised learning and word embeddings.
problem Named Entity Recognition on informal Turkish text types.
method Semi-supervised learning with neural networks and word embeddings.
result Achieved better F-score performances than previous Turkish NER systems.
Study proposes ATS algorithm to solve market clearing problem in Turkish day-ahead market.
problem Market clearing problem in Turkish day-ahead electricity market.
method Proposes an adaptive tabu search (ATS) algorithm to solve the problem, discretizing continuous search space and using adaptive radius.
result ATS algorithm performs better than heuristic decomposition method in most synthetic data sets.
Bidders in day-ahead electricity markets want to sell/buy electricity when their bids generate positive surplus and not to take an action when the reverse holds. However, non-convexities in these markets cause conflicts between the actions that the bidders want to take and the actual market results. In this work, we in…
Simple models outperformed sophisticated ones in forecasting Turkish lira exchange rates.
problem Forecasting Turkish lira exchange rates through univariate techniques.
method Used several models including simple exponential smoothing to predict daily exchange rates.
result Simple exponential smoothing model outperformed all other alternatives.
This paper tackles tweet classification by identifying purpose and position.
problem Difficulties in determining user intention and attitude in short, informal tweets.
method Transformed tweet classification into a multi-label problem and applied a multi-label classification method with post-processing.
result The method effectively classifies tweet purpose and position, outperforming individual classification methods.
Study examines bias in language models across multiple languages.
problem Assessing bias in language models across different languages.
method Semi-automatically translated data sets into multiple languages, analyzed mono- and multilingual models.
result Notable differences in bias across languages, with Turkish models showing least stereotypes.
Study compares shareholder networks in Turkish and Dutch companies.
problem Understanding differences in shareholder roles across countries.
method Constructed shareholder networks, analyzed properties, created randomised versions.
result Roles of different shareholder types differ between Turkey and the Netherlands.
This article presents a preliminary approach towards characterizing political fake news on Twitter through the analysis of their meta-data. In particular, we focus on more than 1.5M tweets collected on the day of the election of Donald Trump as 45th president of the United States of America. We use the meta-data embedd…
Deep autoencoders improve health tweet clustering.
problem Accurate clustering of health-related tweets.
method Deep convolutional autoencoders for learning compact tweet representations.
result Clustering performance significantly outperforms conventional methods.
I-AID categorizes disaster tweets into useful information types.
problem Filtering useful information from social media during disasters.
method Multimodel approach using BERT, GAT, and Relation Network.
result I-AID outperforms state-of-the-art approaches in F1 scores.
In this paper we analyze both the scientific activities of Cahit Arf, a Turkish mathematician, and the social context in which he worked. We also discuss his work and social environment leading to the discovery of Arf invariant, Arf rings, Arf closure and Hasse-Arf theorem.
Analyzes how Trump's tweets impact global stock markets.
problem Understanding the financial impact of presidential tweets on stock markets.
method Examined tweets from Donald Trump's presidency, collected from The Guardian and Bloomberg, and analyzed their effect on equity indices.
result Identified tweets that significantly influenced stock market indices.
ArSentD-LEV dataset improves sentiment analysis in Levantine Arabic tweets.
problem Challenges in sentiment analysis of Arabic tweets, especially Levantine dialect.
method Created a dataset of 4,000 Levantine Arabic tweets with detailed sentiment and topic annotations.
result Improved performance of sentiment classifiers with detailed annotations.
Paper presents a model for identifying informative COVID-19 tweets.
problem Identifying informative COVID-19 tweets on Twitter.
method Leveraged transformers (RoBERTa, XLNet, BERTweet) trained in Semi-Supervised Learning (SSL) setting.
result Achieved F1 score of 0.9011 on test set, ranking 7th on leaderboard.
Study finds mixed evidence of monthly stock market anomalies in Turkey and US.
problem Investigating whether stock markets exhibit abnormal returns monthly.
method Statistical summary analysis, decomposition technique, dummy variable estimation, binary logistic regression.
result Weak evidence against efficient market hypothesis on monthly returns, with notable May effect in Turkey.
Deep learning model extracts location references from tweets during emergencies.
problem Challenges in extracting reliable location information from tweets during crises.
method Convolutional Neural Network (CNN) based model.
result Achieved high accuracy in extracting location references from tweets.
Improved neural model predicts gender from tweets.
problem Predicting gender from Twitter text.
method RNN model with attention, LSA-reduced n-gram features.
result Improved model achieves state-of-the-art performance on English tweets.
Cluster analysis is a field of data analysis that extracts underlying patterns in data. One application of cluster analysis is in text-mining, the analysis of large collections of text to find similarities between documents. We used a collection of about 30,000 tweets extracted from Twitter just before the World Cup st…
Deep learning model classifies tweets for disaster rescue scheduling.
problem Efficiently processing and categorizing tweets for disaster rescue.
method Combining attention-based Bi-directional LSTM and CNN with pre-trained word vectors for classification and feature engineering.
result Proposed model outperforms other methods in Precision, Recall, F1-score, and Accuracy.
Nowcasts consumption and investment from Turkish bank transactions for real-time GDP.
problem Lack of timely data for GDP nowcasting in Turkey.
method Used aggregated bank transactions to replicate quarterly national accounts data.
result Bank transaction data successfully nowcasts Turkish GDP, especially useful for countries with long data lags.
Since the events of the Arab Spring, there has been increased interest in using social media to anticipate social unrest. While efforts have been made toward automated unrest prediction, we focus on filtering the vast volume of tweets to identify tweets relevant to unrest, which can be provided to downstream users for …
Study categorizes and analyzes emotions in sexist tweets.
problem Lack of defined categories for sexism in NLP.
method Used a new dataset from SemEval-2018 to classify and analyze emotions in sexist tweets.
result Demonstrated the mental state and affectual state of users who tweet in different categories of sexism.
D2KLab's approach predicts tweet engagement using two stages.
problem Predicting user engagement with tweets.
method Two-stage approach: feature learning and ensemble XGBoost.
result Ranked 22 in the 2020 RecSys Challenge leaderboard.
Paper proposes Experts Model for better emotion detection in tweets.
problem Estimating intensity of emotion in tweets.
method Inspired by Mixture of Experts (MoE) model, each expert learns different features.
result Our Experts Model stands at top-5 results in emotion detection.
Gang violence is a severe issue in major cities across the U.S. and recent studies [Patton et al. 2017] have found evidence of social media communications that can be linked to such violence in communities with high rates of exposure to gang activity. In this paper we partnered computer scientists with social work rese…
Social networking sites such as Twitter have provided a great opportunity for organizations such as public libraries to disseminate information for public relations purposes. However, there is a need to analyze vast amounts of social media data. This study presents a computational approach to explore the content of twe…
Dataset analyzes tweets' impact on stock returns.
problem Understanding how public opinion affects stock market outcomes.
method Created a dataset of 862,231 labeled tweets, provided baselines and multi-view learning approach.
result Demonstrated the impact of tweets on stock returns over various time frames.
Framework detects cyber threats from Twitter tweets.
problem Time-consuming manual extraction of cyber threat intelligence.
method Novelty detection model trained on CVE data.
result F1-score of 0.643 for classifying cyber threat tweets.
TBIP uses texts to quantify lawmakers' political positions.
problem Quantifying lawmakers' political positions from speeches, tweets, etc.
method Unsupervised probabilistic topic model analyzing texts.
result TBIP separates lawmakers by party and infers ideal points close to vote-based.
A lack of information exists about the health issues of lesbian, gay, bisexual, transgender, and queer (LGBTQ) people who are often excluded from national demographic assessments, health studies, and clinical trials. As a result, medical experts and researchers lack a holistic understanding of the health disparities fa…
New tool detects weak and strong Islamophobic hate speech on social media.
problem Detecting Islamophobic hate speech on social media is challenging due to its varied nature.
method Built a multi-class classifier distinguishing between non-Islamophobic, weak Islamophobic, and strong Islamophobic content using GloVe word embeddings.
result Accuracy of 77.6% and balanced accuracy of 83% on a dataset of 109,488 tweets.
Simple model classifies traffic tweets from Twitter using word embeddings.
problem Sparsity and curse of dimensionality in bag-of-words representation for traffic tweets.
method Proposes a word embedding-based framework to classify traffic-related tweets from non-traffic ones.
result State-of-the-art test accuracy of 95.9% achieved with a simple model.
Crowdsourcing and active learning reduce manual annotation in social media event classification.
problem Manual annotation is time-consuming and resource-intensive in social media event classification.
method Crowdsourcing pipeline combined with active learning strategies.
result Active learning strategies help reduce the number of tweets needed for classification.
This paper studies users' perception regarding a controversial product, namely self-driving (autonomous) cars. To find people's opinion regarding this new technology, we used an annotated Twitter dataset, and extracted the topics in positive and negative tweets using an unsupervised, probabilistic model known as topic …
This paper tackles spam detection on Twitter by analyzing correlated features.
problem Spam detection on social media, especially Twitter, to improve user experience.
method Extracted tweet-based and user-based features, identified correlated features, and used artificial neural networks for classification.
result Achieved 97.57% accuracy in classifying tweets as spam or non-spam.
Improved offensive language detection in tweets with multiple deep learning models.
problem Detecting offensive language in tweets using machine learning.
method Combination of multiple deep learning architectures for classification.
result Achieved macro-average F1-scores of 0.76, 0.68, 0.54 for different tasks.
A dynamic keyword selection model for topic modeling of tweets.
problem Adjusting keywords dynamically to mimic past topics with novelty.
method Generative process selects keywords and documents, trained with variational lower bound and stochastic gradient optimization.
result Keyword-based topic model outperforms a sophisticated baseline model by 67%.
Study predicts cryptocurrency price movements using Twitter sentiment analysis.
problem Predicting short-term price movements of cryptocurrencies.
method Conditional examination of return and excess return rates following tweet publication.
result Statistically significant increases in return rates within the first three minutes after tweet publication.
System detects financial forecasts in tweets, achieving high precision.
problem Detecting financial forecasts in social media messages.
method Natural Language Processing and Machine Learning techniques for real-time analysis.
result Achieves over 90% precision for financial forecasts.
Study uses LLMs to categorize financial tweets, revealing useful sentiment signals.
problem Discovering meaningful sentiment signals from unstructured financial social media data.
method Leveraged LLMs to automatically label financial tweets with event categories and aligned with returns.
result Certain event labels consistently yield negative alpha, with statistically significant Sharpe ratios and information coefficients.
Study examines how social media sentiment impacts biotech stocks.
problem Understanding the impact of social media on biotech stock prices.
method VADER sentiment analysis, ARIMA, and VAR models were used to forecast stock market performance.
result Complex interplay between tweet sentiment and stock market performance was identified.
This paper classifies tweets into positive and negative sentiments using distributed word and sentence representations.
problem Classifying tweets into positive and negative sentiments.
method Used distributed representations of words and sentences, and LSTM and CNN networks for classification.
result Achieved accuracies as high as 81%.
GCAN detects fake news on social media with explanations.
problem Detecting fake news on social media with explanations.
method Graph-aware Co-Attention Networks (GCAN).
result GCAN significantly outperforms state-of-the-art methods in accuracy.
Adversarial tweets can fool stock prediction models, causing financial loss.
problem Vulnerability of stock prediction models to adversarial attacks on social media.
method Solving combinatorial optimization problems with semantic and budget constraints to generate adversarial tweets.
result Adversarial tweets can fool stock prediction models and cause significant financial loss.
Study shows tweets about COVID-19 can predict stock market performance.
problem Understanding the impact of COVID-19 on stock markets.
method Text sentiment analysis of Twitter data to correlate tweets about COVID-19 with stock market performance.
result Strong relationship between COVID-19 sentiment and stock market performance can be predicted.
This paper improves typhoon intensity prediction using social media data and semantic word embeddings.
problem Short-term disaster prediction from historical data alone is limited.
method Combining semantically-enriched word embeddings with traditional word2vec for social media data, and an end-to-end learning framework.
result Our approach outperforms state-of-the-art baselines in typhoon intensity prediction.
Study analyzes global public sentiment on DeFi from 2012-2022.
problem Global public sentiment on DeFi is understudied.
method Sentiment analysis, spatial econometrics, clustering, topic modeling.
result Economic development significantly influences DeFi engagement, especially after 2015.
Study predicts Bitcoin volatility using Twitter data.
problem Forecasting Bitcoin volatility with social media data.
method Deep learning models using Twitter data, including semantic and user statistics.
result Temporal convolutional networks outperform other models in volatility prediction.