Read this BLOG to see how I fine-tuned a sparse transformer on this dataset.

Dataset Description

The Twitter Financial News dataset is an English-language dataset containing an annotated corpus of finance-related tweets. This dataset is used to classify finance-related tweets for their topic.

topics = {
    "LABEL_0": "Analyst Update",
    "LABEL_1": "Fed | Central Banks",
    "LABEL_2": "Company | Product News",
    "LABEL_3": "Treasuries | Corporate Debt",
    "LABEL_4": "Dividend",
    "LABEL_5": "Earnings",
    "LABEL_6": "Energy | Oil",
    "LABEL_7": "Financials",
    "LABEL_8": "Currencies",
    "LABEL_9": "General News | Opinion",
    "LABEL_10": "Gold | Metals | Materials",
    "LABEL_11": "IPO",
    "LABEL_12": "Legal | Regulation",
    "LABEL_13": "M&A | Investments",
    "LABEL_14": "Macro",
    "LABEL_15": "Markets",
    "LABEL_16": "Politics",
    "LABEL_17": "Personnel Change",
    "LABEL_18": "Stock Commentary",
    "LABEL_19": "Stock Movement",
}

The data was collected using the Twitter API. The current dataset supports the multi-class classification task.

Task: Topic Classification

Data Splits

There are 2 splits: train and validation. Below are the statistics:

Licensing Information

The Twitter Financial Dataset (topic) version 1.0.0 is released under the MIT License.