Browse Datasets
Sort by # Views, desc
Turkish Crowdfunding Startups
This dataset contains data on crowdfunding campaigns in Turkey. The dataset includes various characteristics such as crowdfunding projects, project descriptions, targeted and raised funds, campaign durations, and number of backers. Collected in 2022, this dataset provides a valuable resource for researchers who want to understand and analyze the crowdfunding ecosystem in Turkey. In total, there are data from more than 1500 projects on 6 different platforms. The dataset is particularly useful for training natural language processing (NLP) and machine learning models. This dataset is an important reference point for studies on the characteristics of successful crowdfunding campaigns and provides comprehensive information for entrepreneurs, investors and researchers in Turkey.
Turkish User Reviews
This dataset contains Turkish comments made by customers on products (computer, tea machine, head phones, modem, parfume, mobile phone, TV, usb)sold on a website. This dataset created by Asst. Prof. Dr. Ekin Ekinci and Prof. Sevinç İlhan Omurca. Please refer to the study "An alternative word embedding approach for knowledge representation in online consumers’ reviews" when using this dataset.
Bengali Hate Speech Detection Dataset
The dataset can be used for hate speech detection in Bengali social media texts. The dataset is categorized into political, personal, geopolitical, religious, and gender abusive hates that are either directed or generalized towards a specific person, entity, or group. The data and lexicons contain content that is racist, sexist, homophobic, and offensive in many different ways. The dataset is collected and subsequently annotated only for research-related purposes. Besides, authors don't take any liability if some statements contain very offensive and hateful statements that are either directed towards a specific person or entity or generalized towards a group. Therefore, please use it at your risk.
HLS-CMDS: Heart and Lung Sounds Dataset Recorded from a Clinical Manikin using Digital Stethoscope
This dataset contains 535 recordings of heart and lung sounds captured using a digital stethoscope from a clinical manikin, including both individual and mixed recordings of heart and lung sounds; 50 heart sounds, 50 lung sounds, and 145 mixed sounds. For each mixed sound, the corresponding source heart sound (145 recordings) and source lung sound (145 recordings) were also recorded. It includes recordings from different anatomical chest locations, with normal and abnormal sounds. Each recording has been filtered to highlight specific sound types, making it valuable for artificial intelligence (AI) research and applications.
COREVQA
Recently, many benchmarks and datasets have been developed to evaluate Vision-Language Models (VLMs) using visual question answering (VQA) pairs, and models have shown significant accuracy improvements. However, these benchmarks rarely test the model's ability to accurately complete visual entailment, for instance, accepting or refuting a hypothesis based on the image. To address this, we propose COREVQA (Crowd Observations and Reasoning Entailment), a benchmark of 5608 image and synthetically generated true/false statement pairs, with images derived from the CrowdHuman dataset, to provoke visual entailment reasoning on challenging crowded images. Our results show that even the top-performing VLMs achieve accuracy below 80%, with other models performing substantially worse (39.98%-69.95%). This significant performance gap reveals key limitations in VLMs’ ability to reason over certain types of image–question pairs in crowded scenes.
0 to 5 of 5