🧠 ESCODER: Episodic Specificity & Autobiographical Detail Coder

Privacy & Ethics Guarantee: All transcript processing runs 100% locally on your machine via local neural inference. No memory narratives, participant IDs, or study data are ever transmitted across the internet or uploaded to cloud servers.

ESCODER is an automated, dual-engine scoring tool designed for cognitive, clinical, and developmental psychology researchers. It integrates two coding frameworks into a unified desktop dashboard:

• Episodic Detail Decomposition (Levine et al., 2002; Kyung et al., 2016): Parses memory narratives into individual grammatical meaning units, tags internal details ({EVENT}, {TIME}, {PLACE}, {PERCEPTUAL}, {EMOTION}) versus semantic/editorial units ({EXTERNAL}), and displays an annotated transcript for human verification. Powered by a local LLM (Qwen 2.5 via Ollama).
• Automated AMT Specificity Classification (Mistica et al., 2024; Williams & Broadbent, 1986): Generates both granular 5-class categorisation (specific, extended, categoric, omission, associate) and validated binary classification (Specific vs. Non-Specific at $\ge 0.55$) using a fine-tuned BERT transformer classifier (amt-2023-08-01).

📦 Download & Quickstart

Step 1: Install Required Free Engines (One-time only)

Before running ESCODER, ensure your computer has the following installed:

1. R (version 4.2 or later): Download R from CRAN
2. Ollama (for local AI inference): Download Ollama

   • Once Ollama is installed and running, open your computer's Terminal (Mac) or Command Prompt (Windows), enter the following command, and wait for the download to finish:(Note: If working on a laptop with 8 GB RAM or integrated graphics, run ollama pull qwen2.5:7b instead; you can select this lighter model directly in the app's settings sidebar).

     ollama pull qwen2.5:14b
     

Step 2: Download the ESCODER Bundle

📥 Download ESCODER v1.0 (.zip from Zenodo)

Persistent DOI: 10.5281/zenodo.23191424

• Unzip the archive and move the escoder folder to your Documents or research directory.

Step 3: Launch

• Mac users: Double-click Run_ESCODER.command
• Windows users: Double-click Run_ESCODER.bat

Note for Mac users (First-time security check):

If macOS alerts you that the file is from an unidentified developer, right-click (or Control-click) Run_ESCODER.command, select Open, and then confirm by clicking Open in the dialog box. You only need to do this once.

Note for Windows users:

Antivirus software may flag the .bat file as a virus. You can add the folder to your exceptions/turn off your antivirus, or simply open app.R in RStudio and click Run App to launch ESCODER.

On first launch, the script checks dependencies, sets up the local Python virtual environment (.venv), verifies Ollama, and opens the dashboard in your default browser.