Uploaded image for project: 'Machine Learning Library'
  1. Machine Learning Library
  2. ML-457

Applying HPCC Systems Word Vectors to SEC Filings

    Details

    • Type: New Feature
    • Status: Open
    • Priority: Not specified
    • Resolution: Unresolved
    • Affects Version/s: None
    • Fix Version/s: None
    • Component/s: None
    • Labels:

      Description

      Deliverables: Will be implemented

      • REPORT on current state of Vectorization and NLP representation of SEC Filings
      • Compilation of identified SEC filing cases and their intersection with Lexis Nexis’ focus
      • FUNCTION sorting SEC data into common SEC filing cases (by form num, etc.)
      • FUNCTION to convert SEC filings data into format required by HPCC Systems Word Vectors, including any FILTERS and TRANSFORMATIONS
      • Transform reformatted SEC data using HPCC Systems Word Vectors and the filing label obtained by previous functions
      • FUNCTION combining Word Vectors with other info (original data, form no., etc.) to create semi-structured data point about the filing
      • EDA/REPORT on sample transformed into unstructured form (number of fields, pct NA, correlations and time series, etc. within and across form nos.)
      • TEST of consistency and latency of transformations
      • Documentation

      Wishlist:

      • Process complete historical SEC filing data for upload by entity (company, officer, etc.)
      • Prepare processed historical data for open-source query
      • FUNCTION to extract sentiment from earnings reports
      • ANALYSIS of share disposition/acquisition over time w.r.t. financials
      • Compile factored analyst responses (positive/negative/neutral) to earnings reports

      FUNCTION to predict analyst sentiment from earnings reports

        Attachments

          Activity

            People

            • Assignee:
              mmurray Matthias Murray
              Reporter:
              lorraineachapman Lorraine Chapman
            • Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

              Dates

              • Created:
                Updated: