CoolFace
Datasetpublic

sohomghosh/MiMIC_Multi-Modal_Indian_Earnings_Calls_Dataset

MiMIC_analysis_code.ipynb : This is the actual code file with results MiMIC_Multi-Modal_Indian_Earnings_Calls.xlsx : This is the main file having to be used for training. Descripton of columns of this file is given below. getting_all_texts_together_embeddings_dim128_CPU.pkl : This is the file having embeddings of texts (extracted from transcripts, images) and tables (extracted from images) concatenated one after the other.… See the full description on the dataset page: https://huggingface.co/datasets/sohomghosh/MiMIC_Multi-Modal_Indian_Earnings_Calls_Dataset.

sourceHugging Facecc-by-nc-sa-4.0updated 1y agoView on Hugging Face
3likes122downloads
Dataset Card

MiMICanalysiscode.ipynb : This is the actual code file with results

MiMICMulti-ModalIndianEarningsCalls.xlsx : This is the main file having to be used for training. Descripton of columns of this file is given below.

gettingalltextstogetherembeddingsdim128CPU.pkl : This is the file having embeddings of texts (extracted from transcripts, images) and tables (extracted from images) concatenated one after the other.

gettingalltexts_together.pkl : This is the file having concatenated texts mentioned above.

imageembeddingmeanpooled128dimCPU_df.csv : This is the file having embeddings of images. For instances with multiple images, we have performed mean pooling.

rawdata/extractedtexts.zip : This is the file having texts extracted from the transcripts and presentations.

rawdata/extractedtablesfromimagesnewfiltered.zip : This is the file having tables extracted from the presentations.

rawdata/extractedimagesnewfiltered.zip : This is the file having images extracted from the presentations.

final_train.csv : Final training file with text & image classifier probabilities

final_valid.csv : Final validation file with text & image classifier probabilities

final_test.csv : Final test file with text & image classifier probabilities

BestPerformingModelonMiMIC.zip : Compressed version of best performing artifact

Description of Columns ----------------------

column_namecolumn_description
company_nameName of the company
tickerTicker symbol of the company
companywebsitelinkLink to the company website
screener_linkLink to the screener
transcript_linkLink to the transcript
transcriptfilenameFile name of the transcript
ppt_linkLink to the presentation
pptfilenameFile name of the presentation
RESULT DATEResult date
yearYear from Result date
month_numericMonth from Result date
RESULT DATE open priceOpen price on the result date
RESULT DATE+1Date following the result date
SMA20Simple Moving Average of 20 periods
SMA50Simple Moving Average of 50 periods
RSI14Relative Strength Index
gdp_growthGDP growth rate
inflationInflation rate
NIFTY50_openNIFTY 50 open price
NIFTY50_closeNIFTY 50 close price
NIFTY50_volumeNIFTY 50 traded volume
previous_yearPrevious year
financial_yearFinancial year
SalesCompany Sales
ExpensesExpenses incurred by company
Operating ProfitOperating Profit
OPM %Operating profit margin
Other IncomeOther income sources
InterestInterest expense
DepreciationDepreciation expense
Profit before taxProfit before tax deduction
Tax %Tax Percentage
Net ProfitNet Profit after tax
EPS in RsEarnings per share
Dividend Payout %Dividend payout ratio
Equity CapitalEquity capital
ReservesReserves held by the company
BorrowingsBorrowings amount
Other LiabilitiesOther liabilities
Total LiabilitiesTotal liabilities of company
Fixed AssetsFixed assets of company
CWIPCapital Work in Progress
InvestmentsInvestments held
Other AssetsOther assets
Total AssetsTotal assets of company
Cash from Operating ActivityCash flow from operating activities
Cash from Investing ActivityCash flow from investing activities
Cash from Financing ActivityCash flow from financing activities
Net Cash FlowNet cash flow
RevenueRevenue generated
Financing ProfitProfit from financing activities
Financing Margin %Financing Margin percentage
DepositsDeposits amount
BorrowingBorrowing amount
RESULT DATE+1 open priceOpen price on the next day
TARGET-2 REGRESSION NORMALIZEDTarget variable for regression (normalized)
splitData split (train, validation, test)
bibtex
@misc{ghosh2025mimicmultimodalindianearnings,
      title={MiMIC: Multi-Modal Indian Earnings Calls Dataset to Predict Stock Prices}, 
      author={Sohom Ghosh and Arnab Maji and Sudip Kumar Naskar},
      year={2025},
      eprint={2504.09257},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2504.09257}, 
}