Context

WhatsApp Messenger, or simply WhatsApp, is an internationally available American freeware, owned by Meta Platforms (previously Facebook). This dataset provides the latest statistics of Whatsapp in our day-to-day lives.

Content

The dataset contains 7 files: * age_group.csv : Whatsapp usage by Age group (US) * by_country.csv : Whatsapp users by country * messages_sent_daily.csv : Whatsapp messages sent daily * ratings.csv : Whatsapp Play Store & App Store ratings * usage.csv: Whatsapp daily, weekly & monthly usage (US) * user.csv: Whatsapp users growth over time * user_growth : Latest Whatsapp users growth percentage

Acknowledgements

This data has been scraped from Bussiness Insider, Twitter, Facebook, Statista, Sensor Tower, backlinkto and some others.

Inspiration

This dataset can be analyzed to: * see the effect of Whatsapp on the present day world; * how much time does an average person spends on Whatsapp; * the number of users on the platform; and a lot other parameters that we can think of!

In 2021, WhatsApp's user base in India amounts to approximately ****** million users. The number of WhatsApp users in India is projected to reach ****** million users by 2025. User figures have been estimated by taking into account company filings or press material, secondary research, app downloads and traffic data. They refer to the average monthly active users over the period.The shown data are an excerpt of Statista's Key Market Indicators (KMI). The KMI are a collection of primary and secondary indicators on the macro-economic, demographic and technological environment in up to *** countries and regions worldwide. All indicators are sourced from international and national statistical offices, trade associations and the trade press and they are processed to generate comparable data sets (see supplementary notes under details for more information).

Context

https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F1842206%2F7394017546a1babbdf7394aa61da2e80%2Fwhatsapp2.png?generation=1700227405061793&alt=media" alt="">

Based on their wikipedia page

WhatsApp (officially WhatsApp Messenger) is a freeware, cross-platform, centralized instant messaging (IM) and voice-over-IP (VoIP) service owned by United States tech conglomerate Meta Platforms. It allows users to send text, voice messages and video messages, make voice and video calls, and share images, documents, user locations, and other content. WhatsApp's client application runs on mobile devices, and can be accessed from computers.The service requires a cellular mobile telephone number to sign up. In January 2018, WhatsApp released a standalone business app called WhatsApp Business which can communicate with the standard WhatsApp client.

These reviews were extracted from Google Store.

Usage

This dataset should paint a good picture on what is the public's perception of the app over the years. Using this dataset, we can do the following

1. Extract sentiments and trends
2. Identify which version of the app had the most positive feedback, the worst.
3. Use topic modeling to identify the pain points of the application.

(AND MANY MORE!)

Note

Images generated using Bing Image Generator

Description:

The "Daily Social Media Active Users" dataset provides a comprehensive and dynamic look into the digital presence and activity of global users across major social media platforms. The data was generated to simulate real-world usage patterns for 13 popular platforms, including Facebook, YouTube, WhatsApp, Instagram, WeChat, TikTok, Telegram, Snapchat, X (formerly Twitter), Pinterest, Reddit, Threads, LinkedIn, and Quora. This dataset contains 10,000 rows and includes several key fields that offer insights into user demographics, engagement, and usage habits.

Dataset Breakdown:

• Platform: The name of the social media platform where the user activity is tracked. It includes globally recognized platforms, such as Facebook, YouTube, and TikTok, that are known for their large, active user bases.

• Owner: The company or entity that owns and operates the platform. Examples include Meta for Facebook, Instagram, and WhatsApp, Google for YouTube, and ByteDance for TikTok.

• Primary Usage: This category identifies the primary function of each platform. Social media platforms differ in their primary usage, whether it's for social networking, messaging, multimedia sharing, professional networking, or more.

• Country: The geographical region where the user is located. The dataset simulates global coverage, showcasing users from diverse locations and regions. It helps in understanding how user behavior varies across different countries.

• Daily Time Spent (min): This field tracks how much time a user spends on a given platform on a daily basis, expressed in minutes. Time spent data is critical for understanding user engagement levels and the popularity of specific platforms.

• Verified Account: Indicates whether the user has a verified account. This feature mimics real-world patterns where verified users (often public figures, businesses, or influencers) have enhanced status on social media platforms.

• Date Joined: The date when the user registered or started using the platform. This data simulates user account history and can provide insights into user retention trends or platform growth over time.

Context and Use Cases:

• This synthetic dataset is designed to offer a privacy-friendly alternative for analytics, research, and machine learning purposes. Given the complexities and privacy concerns around using real user data, especially in the context of social media, this dataset offers a clean and secure way to develop, test, and fine-tune applications, models, and algorithms without the risks of handling sensitive or personal information.

Researchers, data scientists, and developers can use this dataset to:

• Model User Behavior: By analyzing patterns in daily time spent, verified status, and country of origin, users can model and predict social media engagement behavior.

• Test Analytics Tools: Social media monitoring and analytics platforms can use this dataset to simulate user activity and optimize their tools for engagement tracking, reporting, and visualization.

• Train Machine Learning Algorithms: The dataset can be used to train models for various tasks like user segmentation, recommendation systems, or churn prediction based on engagement metrics.

• Create Dashboards: This dataset can serve as the foundation for creating user-friendly dashboards that visualize user trends, platform comparisons, and engagement patterns across the globe.

• Conduct Market Research: Business intelligence teams can use the data to understand how various demographics use social media, offering valuable insights into the most engaged regions, platform preferences, and usage behaviors.

• Sources of Inspiration: This dataset is inspired by public data from industry reports, such as those from Statista, DataReportal, and other market research platforms. These sources provide insights into the global user base and usage statistics of popular social media platforms. The synthetic nature of this dataset allows for the use of realistic engagement metrics without violating any privacy concerns, making it an ideal tool for educational, analytical, and research purposes.

The structure and design of the dataset are based on real-world usage patterns and aim to represent a variety of users from different backgrounds, countries, and activity levels. This diversity makes it an ideal candidate for testing data-driven solutions and exploring social media trends.

Future Considerations:

As the social media landscape continues to evolve, this dataset can be updated or extended to include new platforms, engagement metrics, or user behaviors. Future iterations may incorporate features like post frequency, follower counts, engagement rates (likes, comments, shares), or even sentiment analysis from user-generated content.

By leveraging this dataset, analysts and data scientists can create better, more effective strategies ...

Dataset Overview: This meticulously curated dataset amalgamates user reviews of the WhatsApp Business application from the App Stores of three distinct nations: Australia, Canada, and the United States. Encompassing a wealth of insights, the dataset is an invaluable resource for data scientists aiming to delve into consumer feedback, analyze sentiment trends, and extract meaningful patterns relevant to business strategies and app improvements. The dataset brings together a comprehensive collection of user interactions and sentiments, shedding light on various facets of the user experience across different geographic locales. With data ethically mined and responsibly aggregated, it ensures a respectful approach towards user privacy while offering rich analytical opportunities.

Data Science Applications: This dataset serves as a fertile ground for a multitude of data science endeavors including sentiment analysis, trend forecasting, natural language processing, and customer feedback analytics. Researchers and practitioners can employ this dataset to harness predictive insights, gauge app performance, and refine user engagement strategies.

Column Descriptors: - ID: Unique identifier for each review entry. - Date: Timestamp of the review. - UserName (Hashed): Anonymized username of the reviewer. - UserUrl (Hashed): Anonymized URL of the user's review profile. - Version: App version the review pertains to. - Score: User-provided rating for the app. - Title: Title of the review. - Text: Body text of the review, providing detailed feedback. - URL: URL of the review on the App Store (anonymized if containing personal identifiers). - Country: The country from which the review was posted. - AppId: Unique identifier for the WhatsApp Business application.

Ethical Considerations: In the assembly of this dataset, utmost care was taken to ensure the ethical sourcing of data, with a keen focus on maintaining the anonymity of the users. Personal identifiers such as usernames and user URLs have been hashed to uphold privacy and confidentiality. SHA-256 was used.

Acknowledgements: Gratitude is extended to the platforms that facilitated the aggregation of this data, acknowledging their pivotal role in fostering a space where user feedback can thrive and inform better business and technological outcomes.

Nigeria whatsapp number list will make your marketing more successful. Also, this confirmed dataset is perfect for making calls or sending messages. Similarly, you can reach people in different parts of Nigeria. Our Nigeria whatsapp number list has over 95% accurate and updated contact numbers. We affirm all the numbers to make sure they are the latest and collect them in an Excel format. If you want to expand your business easily, this Nigeria whatsapp number leads is just right for you. To that end, people can take it from us without any doubt. Nigeria whatsapp phone number data can help you connect with many people and grow your business. In other words, this List To Data is ideal for reaching potential customers all across the country. Besides, you can always use it for direct and online marketing. This is also an easy way to talk directly and interact with interested people. Thus, the Nigeria whatsapp phone number data is an excellent tool for business in Nigeria. In fact, there are over 51 million whatsapp users in Nigeria. With this contact dataset, you can easily follow your marketing plans. Most importantly, through this library, you can earn a great return on investment (ROI).

Don't forget to upvote, comment, and follow if you are using this dataset. If you have any questions about the dataset I uploaded, feel free to leave them in the comments. Thank you! :)

Jangan lupa untuk upvote, comment, follow jika anda menggunakan dataset ini, dan jika ada pertanyaan mengenai dataset yang saya upload, silahkan tinggalkan di comment. Terima kasih :)

Column Descriptions (English) 1. reviewId: A unique ID for each user review. 2. userName: The name of the user who submitted the review. 3. userImage: The URL of the user's profile picture. 4. content: The text content of the review provided by the user. 5. score: The review score given by the user, typically on a scale of 1-5. 6. thumbsUpCount: The number of likes (thumbs up) received by the review. 7. reviewCreatedVersion: The app version used by the user when creating the review (not always available). 8. at: The date and time when the review was submitted. 9. replyContent: The developer's response to the review (no data available in this column). 10. repliedAt: The date and time when the developer's response was submitted (no data available in this column). 11. appVersion: The app version used by the user when submitting the review (not always available).

Deskripsi Kolom (Bahasa Indonesia) 1. reviewId: ID unik untuk setiap ulasan yang diberikan pengguna. 2. userName: Nama pengguna yang memberikan ulasan. 3. userImage: URL gambar profil pengguna yang memberikan ulasan. 4. content: Isi teks ulasan yang diberikan oleh pengguna. 5. score: Skor ulasan yang diberikan pengguna, biasanya dalam skala 1-5. 6. thumbsUpCount: Jumlah suka (thumbs up) yang diterima oleh ulasan tersebut. 7. reviewCreatedVersion: Versi aplikasi yang digunakan pengguna saat membuat ulasan (tidak selalu tersedia). 8. at: Tanggal dan waktu saat ulasan dibuat. 9. replyContent: Isi balasan dari pengembang aplikasi terhadap ulasan (tidak ada data dalam kolom ini). 10. repliedAt: Tanggal dan waktu saat balasan dari pengembang diberikan (tidak ada data dalam kolom ini). 11. appVersion: Versi aplikasi yang digunakan pengguna saat memberikan ulasan (tidak selalu tersedia).

Italy whatsapp number list will make your marketing more successful. In other words, this verified data is perfect for making calls or sending messages. Likewise, you can reach people in different parts of Italy. Our Italy whatsapp number list has over 95% accurate and up-to-date contact numbers. We verify all the numbers to make sure they are real and reliable and collect them in an Excel format. If you want to expand your business easily, this Italy whatsapp number list is just right for you. Our team checks the dataset several times and discards any false leads. Indeed, Everybody can take it from us without any doubt. Italy whatsapp phone number data can help you connect with many people and grow your business. Also, this List To Data is perfect for reaching potential customers all across the country. Moreover, you can always use it for direct and digital marketing. This is also an easy way to talk directly to interested people. Thus, the Italy whatsapp phone number data is an excellent tool for business in Italy. Right now, there are over 37.3 million whatsapp users in Italy. With our trustworthy contact dataset, you can easily follow your marketing plans. Most importantly, through this directory, you can enjoy a great return on investment (ROI).

Japan whatsapp number list makes your business run smoothly. This contact number lead is perfect for your digital marketing campaigns. Japan whatsapp number list is quick and easy to use. Many people in Japan check their phones regularly for calls and messages. This database was created by following the full GDPR policy. People in Japan stay connected and are always looking for new offers. So, it supports your efforts to boost your customer base and marketing site. Likewise, List To Data grows your business profit and helps you get better results. So, buy this and grow your profits and earn a massive return on investment [ROI]. Japan whatsapp phone number data can help your company grow. Most importantly, it supplies a great tool to connect with your audience. Also, the dataset includes the most 95% authentic and valuable whatsapp numbers available. Using this contact list, you can achieve your goal very easily. The Japan whatsapp phone number data supports many people to connect with your business. If you take these special offers or updates, this contact directory will help you connect with more customers. Once you place your order and pay, we will send you this dataset in a CSV format. Mostly, we need full payment in advance to process the order.

Lebanon whatsapp number list will make your marketing more successful. In other words, this verified lead is excellent for sending messages about the product. Likewise, you can reach people in different parts of Lebanon. Our Lebanon whatsapp number list has over 95% accurate and up-to-date contact numbers. We verify all the numbers to make sure they are real and trustworthy and collect them in an Excel format. If you want to expand your business easily, this Lebanon whatsapp number leads is just right for you. Overall, the special team checks the dataset several times and discards any false leads. Indeed, people can accept it from us without any doubt. Lebanon whatsapp phone number data can assist you connect with many people and grow your business. In fact, this List To Data is perfect for reaching potential clients all across the country. Moreover, anyone can always utilize it for direct and digital marketing. This is also an easy path to talk directly to interested people. Thus, the Lebanon whatsapp phone number data is an excellent tool for business in Lebanon. Right now, there are over 2.1 million whatsapp users in Lebanon. Most importantly, through this contact database, you can enjoy a great return on investment (ROI).

This dataset predominantly features user-generated reviews and ratings of the Whatssap Android App, with updates made on a daily basis. It also provides details on the relevancy of the reviews and the dates when they were uploaded.

Luxembourg whatsapp number list is a potential contact data for B2B and B2C marketing campaigns. Again, the whatsapp number lead also lets you track how well your marketing is working. If any dealer or seller wants to promote their business in online marketing, they can contact us. Thus, our consumer database is up-to-date. So, the Luxembourg whatsapp number list helps to connect with many renowned business owners. We give this number dataset in a CSV or an Excel format. Most importantly, that can be downloadable on any CRM platform. To that end, to get the best result from business, you must rely on us. Luxembourg whatsapp phone number data helps to connect many reliable customers for digital marketing. Firstly, our List To Data website is very secure and we give you 95% exact contact numbers. Approximately, 673 thousand people live here and most of them are Internet users. In addition, it can be a potential country for running any online business. Businessmen can easily fulfill their dreams via it. Mainly, we deliver you this service at the most satisfactory price. From this, people will gain a remarkable return on investment (ROI). Yet, the experts create Luxembourg whatsapp phone number data and maintain the GDPR laws strictly for safety.

High Frequency Indicator: The dataset contains year- and month-wise compiled data from the year 2021 to till date on the number of different types of grievances (complaints) received from the users by WhatsApp and the action taken by it. The data compiled is based on the monthly transparency reports published by WhatsApp in accordance with Rule 4(1)(d) of the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 (IT Rules, 2021).

The types of grievances received include Account Support, Safety, Ban Appeal, Other Support and Product Support. The action taken includes the accounts actioned on the basis of grievances received.

Notes:

1. ‘Accounts Actioned’ denotes reports where WhatsApp took remedial action based on the report. Taking action denotes either banning an account or a previously banned account being restored as a result of the complaint.
2. ‘Other’ support pertains to requests which are not consistently classifiable
3. ‘N/A’ denotes grievance topics where it’s generally not applicable to action an account. In some rare cases, tickets may be later reassigned to another topic, and an action could arise from the re-assignment.

The number of WhatsApp users in Brazil was forecast to continuously increase between 2024 and 2029 by in total *********** users (***** percent). After the ninth consecutive increasing year, the WhatsApp user base is estimated to reach ****** million users and therefore a new peak in 2029. Notably, the number of WhatsApp users of was continuously increasing over the past years.User figures, shown here regarding the platform whatsapp, have been estimated by taking into account company filings or press material, secondary research, app downloads and traffic data. They refer to the average monthly active users over the period.The shown data are an excerpt of Statista's Key Market Indicators (KMI). The KMI are a collection of primary and secondary indicators on the macro-economic, demographic and technological environment in up to *** countries and regions worldwide. All indicators are sourced from international and national statistical offices, trade associations and the trade press and they are processed to generate comparable data sets (see supplementary notes under details for more information).

This dataset presents year and month wise enforcement actions taken by Significant Social Media Intermediaries (SSMIs) from 2021 to the present, compiled from the mandatory monthly transparency reports published under Rule 4(1)(d) of the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021. It includes counts of content removed, accounts suspended or banned, and chatrooms, comments, edit profiles and livestreams restricted, along with the policy or violation category (e.g., child sexual exploitation, terrorism, hate speech, bullying, violence, regulated goods, misinformation, etc.).

To enable comparability across platforms with different reporting terms, the dataset uses a standardised enforcement classification:

1. enforcement_type:

The type of action taken: a. Content Actioned (any enforcement such as warning, downranking, age-gating), b. Content Removed (content deleted or made inaccessible), c. Account Banned (account suspension or disabling), d. Quality Metric (AI moderation accuracy indicators reported by some platforms).

1. proactive_flag:

Whether the platform identified and enforced before user reports: a. Proactive = Found via automated detection or internal review systems, b. Unknown = Platform did not specify proactive vs reactive.

Notes: 1. SSMI denotes to Significant Social Media Intermediaries, with over 50,00,000 registered users in India, which primarily or solely enables online interaction between two or more users and allows them to create, upload, share, disseminate, modify or access information using its services

1. Facebook & Instagram (Meta) a. Content Actioned counts any enforcement, not only removals (e.g., removals, warning screens/covering, age gates, downranking). b. Proactive Rate = (items found & actioned proactively) ÷ (total content actioned).

2. X/Twitter a. Child Sexual Exploitation and terrorism suspensions are largely proactive, flagged using proprietary tools and industry hash-sharing systems. b. Data reflects global enforcement, not only India.

3. Google / YouTube a. Number of removal actions as a result of automated detection captures actions triggered by automated systems (ML + human-trained models).

4. ShareChat a. Content Removed / Taken Down / UGC discard / Comments/Chatrooms deleted are standardised as Content Removed. b. Also includes rights-holder reporting workflow for copyright/IP and automated proactive monitoring for harmful content.

5. WhatsApp a. Reports Proactively Banned Accounts, meaning accounts banned before any user reports.

6. Koo a. Distinguishes between Content Removed, Content Actioned (flagged/downranked), and Account Banned. b. Automation Correct/Wrong reflect AI moderation accuracy, not enforcement outcomes.

Ivory Coast whatsapp number list provides you with real contact numbers for active whatsapp users. These users are ready to engage with your business and receive your marketing messages. With this data, you can easily connect with these users, promote your brand, and grow your customer base. Most importantly, the contact numbers in our database are reliable and accurate. You can trust that the contacts are real and up-to-date because our experts check them regularly. You can send special offers, updates, and information directly to interested customers. Buy valuable contact data from List to Data at wholesale prices. Ivory Coast whatsapp phone number data is a valuable tool for your business because it helps you quickly connect with many people. With this data, you can reach customers right on their phones. Ivory Coast whatsapp contact number database will help you find the right audience. You can also share news about your products and services with just a few taps. As a result, this makes it faster than sending emails or using other methods. On the other hand, people in your country like using whatsapp because it is simple and quick. Therefore, when you use this data, you can talk to customers who want to hear from you.

UK whatsapp number list is one of the latest tools that will meet your needs. Moreover, this directory is the latest creation of List to Data. With this, you can contact the corporate leaders of your target firms, who might assist you in achieving your objective. Furthermore, we keep the UK whatsapp number list GDPR standards and have well-constructed verified data. We regularly update the data as well. Not to mention that it will allow you to interact with potential investors. It will therefore assist you in generating more sales leads and boosting your company’s revenue. UK whatsapp phone number data contains the contact addresses of the people who live in the country. You can get thousands of contacts and other information in this database. UK whatsapp phone number data can boost digital marketing campaigns. Running an online campaign is crucial right now if you want to stay in the spotlight. You cannot accomplish anything without the internet and digital marketing. Indeed, we all understand that at this time it is a must to communicate. We cannot overlook the fact that gadgets are used by everyone. Because of this, we must face reality and update our marketing strategy.

Dataset

This dataset was created by Hemant Palivela

Released under Data files © Original Authors

Contents

Singapore whatsapp number list helps businesses connect easily with their audience. You can sort the information by gender, age, relationship status, and place. This filtering helps you easily target specific groups who are more likely to have an interest in your product or service. Singapore whatsapp data is valid, which means you always reach active and real users. If you ever need help, 24/7 support is available. This means someone is always ready to assist you with any issues you may face, so you never feel alone. Many businesses use the List to Data to grow and reach more customers. Singapore whatsapp phone number data is a collection of phone numbers that businesses use to reach customers in Singapore. All the numbers in this database are GDPR-protected. This means your data stays safe and follows privacy laws. Singapore whatsapp phone number data helps you reach the right people quickly. You can relax because there are no rules to break. Also, the database has a replacement guarantee. If any number doesn’t work, you will get a new one quickly. This feature keeps your list fresh and reliable. It ensures that you always connect with active users who want to hear from you.

MultiSocial is a dataset (described in a paper) for multilingual (22 languages) machine-generated text detection benchmark in social-media domain (5 platforms). It contains 472,097 texts, of which about 58k are human-written and approximately the same amount is generated by each of 7 multilingual large language models by using 3 iterations of paraphrasing. The dataset has been anonymized to minimize amount of sensitive data by hiding email addresses, usernames, and phone numbers.

If you use this dataset in any publication, project, tool or in any other form, please, cite the paper.

Disclaimer

Due to data source (described below), the dataset may contain harmful, disinformation, or offensive content. Based on a multilingual toxicity detector, about 8% of the text samples are probably toxic (from 5% in WhatsApp to 10% in Twitter). Although we have used data sources of older date (lower probability to include machine-generated texts), the labeling (of human-written text) might not be 100% accurate. The anonymization procedure might not successfully hiden all the sensitive/personal content; thus, use the data cautiously (if feeling affected by such content, report the found issues in this regard to dpo[at]kinit.sk). The intended use if for non-commercial research purpose only.

Data Source

The human-written part consists of a pseudo-randomly selected subset of social media posts from 6 publicly available datasets:

1. Telegram data originated in Pushshift Telegram, containing 317M messages (Baumgartner et al., 2020). It contains messages from 27k+ channels. The collection started with a set of right-wing extremist and cryptocurrency channels (about 300 in total) and was expanded based on occurrence of forwarded messages from other channels. In the end, it thus contains a wide variety of topics and societal movements reflecting the data collection time.

2. Twitter data originated in CLEF2022-CheckThat! Task 1, containing 34k tweets on COVID-19 and politics (Nakov et al., 2022, combined with Sentiment140, containing 1.6M tweets on various topics (Go et al., 2009).

3. Gab data originated in the dataset containing 22M posts from Gab social network. The authors of the dataset (Zannettou et al., 2018) found out that “Gab is predominantly used for the dissemination and discussion of news and world events, and that it attracts alt-right users, conspiracy theorists, and other trolls.” They also found out that hate speech is much more prevalent there compared to Twitter, but lower than 4chan's Politically Incorrect board.

4. Discord data originated in Discord-Data, containing 51M messages. This is a long-context, anonymized, clean, multi-turn and single-turn conversational dataset based on Discord data scraped from a large variety of servers, big and small. According to the dataset authors, it contains around 0.1% of potentially toxic comments (based on the applied heuristic/classifier).

5. WhatsApp data originated in whatsapp-public-groups, containing 300k messages (Garimella & Tyson, 2018). The public dataset contains the anonymised data, collected for around 5 months from around 178 groups. Original messages were made available to us on request to dataset authors for research purposes.

From these datasets, we have pseudo-randomly sampled up to 1300 texts (up to 300 for test split and the remaining up to 1000 for train split if available) for each of the selected 22 languages (using a combination of automated approaches to detect the language) and platform. This process resulted in 61,592 human-written texts, which were further filtered out based on occurrence of some characters or their length, resulting in about 58k human-written texts.

The machine-generated part contains texts generated by 7 LLMs (Aya-101, Gemini-1.0-pro, GPT-3.5-Turbo-0125, Mistral-7B-Instruct-v0.2, opt-iml-max-30b, v5-Eagle-7B-HF, vicuna-13b). All these models were self-hosted except for GPT and Gemini, where we used the publicly available APIs. We generated the texts using 3 paraphrases of the original human-written data and then preprocessed the generated texts (filtered out cases when the generation obviously failed).

The dataset has the following fields:

• 'text' - a text sample,

• 'label' - 0 for human-written text, 1 for machine-generated text,

• 'multi_label' - a string representing a large language model that generated the text or the string "human" representing a human-written text,

• 'split' - a string identifying train or test split of the dataset for the purpose of training and evaluation respectively,

• 'language' - the ISO 639-1 language code identifying the detected language of the given text,

• 'length' - word count of the given text,

• 'source' - a string identifying the source dataset / platform of the given text,

• 'potential_noise' - 0 for text without identified noise, 1 for text with potential noise.

ToDo Statistics (under construction)