GUI_tutorial.md

Graphical User Interface Tutorial

Translated from Chinese tutorials by GPT-4, refer to the Chinese tutorials for accuracy.

1. Prepare API Keys (This tutorial is from GPT-4)

The GUI running license is located in the License file in the root folder of this repository. Place it in the same folder as the exe file. Please do not modify the filename of the License file.

Preparing API keys is a prerequisite for using various API services. This guide covers how to obtain API keys for SerpAPI, LLM,和Elsevier.

SerpAPI Key

1. Register or Login
   • Visit the SerpAPI website, register an account or login if you already have one.
2. Access Dashboard
   • After logging in, navigate to your user dashboard.
3. Find or Generate API Key
   • On the dashboard, you should be able to see your API key. If not, there might be an option to generate a new key.
4. Copy API Key
   • Copy the API key for use in your application.

LLM Key (Supports Claude2 API and OpenAI format API)

1. Register or Login
   • Identify the LLM (Large Language Model) API provider (such as OpenAI), create an account or login.
2. Acquire API Access
   • Go to the website's API or developer section.
3. Generate API Key
   • Look for an option to generate a new API key and follow the prompts to create one.
4. Copy and Secure Key
   • Copy the generated API key and store it securely, as it allows access to LLM services.

Elsevier Key

1. Elsevier Account
   • Visit Elsevier's Developer Portal, register or login.
2. Create New API Key
   • Navigate to the section where you can create or manage API keys. This might be in your account settings or a specific "API Keys" section.
3. Application Details
   • You might need to provide details about your application, including its name and purpose, to generate an API key.
4. Obtain and Store Key
   • After submission, your API key will be displayed. Copy this key and keep it securely stored.

General Tips

• Security: Keep your API keys confidential to prevent unauthorized access.
• Regeneration: If a key is compromised, regenerate it from the service's dashboard as soon as possible.
• Usage Limits: Be aware of any usage limits or quotas associated with your API keys to avoid service interruptions.

2. Set Environment Variables (This tutorial is from GPT-4)

The environment variable to set is: ElsClientKey, which is the Elsevier key obtained in the previous step.

In Windows 10, you can use command line tools, such as cmd or PowerShell, to set environment variables. Here is a guide to the two main methods.

Via Command Prompt (CMD)

1. Set Temporary Environment Variable:

   • Open Command Prompt (CMD).
   • Use the set command to set an environment variable. This sets a temporary variable that disappears after closing the CMD window. For example, to set an environment variable named MYVAR with the value value:

     set MYVAR=value
     

2. Set Permanent Environment Variable:

   • For a permanent environment variable, you need to use the setx command. This permanently adds or modifies an environment variable, but only affects new command line windows. For example:

     setx MYVAR "value"
     

   • To set an environment variable for all users, use the /M switch:

     setx /M MYVAR "value"
     

   Note that when using setx, if the environment variable value contains spaces, it needs to be enclosed in quotes.

Via PowerShell

1. Set Temporary Environment Variable:

   • Open PowerShell.
   • Use $env: to set a temporary environment variable, which will be effective in the current PowerShell session. For example:

     $env:MYVAR = "value"
     

2. Set Permanent Environment Variable:

   • Use the [System.Environment]::SetEnvironmentVariable method. This allows you to set a permanent environment variable for the current user or all users. For example, to set for the current user:

     [System.Environment]::SetEnvironmentVariable("MYVAR", "value", [System.EnvironmentVariableTarget]::User)
     

   • To set for all users, make sure to run PowerShell as an administrator, then execute:

     [System.Environment]::SetEnvironmentVariable("MYVAR", "value", [System.EnvironmentVariableTarget]::Machine)
     

   This requires administrative privileges.

Notes

• When using the setx command, the maximum character length limit is 1024 characters.
• After setting an environment variable, you may need to restart your command line tool or computer for the changes to take effect.
• When setting environment variables with PowerShell, choose User or Machine as the target, representing the current user and all users, respectively.

3. Download and Run the Pre-packaged exe File (This tutorial is from GPT-4)

The following steps will guide you on how to download and run an exe file from a specified GitHub page.

Steps

1. Access the GitHub Release Page
   • Open a browser and visit the AutomaticReviewGeneration release page.
2. Select a Version
   • Browse through the different release versions and select the one you need to download.
3. Download the exe File
   • Find the file ending in .exe in the selected version. Click the download link next to the file name to download.
4. Run the exe File
   • After downloading, locate the downloaded .exe file and double-click to run.
   • If the system prompts "Unknown publisher," choose "Run" to continue.

4. Interface Button Introduction

The interface after opening the program is as follows

• TOPIC: The topic of the generated review
• Demo: Whether to generate a test with a small amount of literature
• Whole Process: Whether to proceed with the whole process, or just perform literature retrieval and download

• Skip Literature Search: Whether to skip the literature search module and proceed to subsequent steps
• Skip Topic Formulation: Whether to skip the topic generation module and proceed to subsequent steps
• Skip Knowledge Extraction: Whether to skip the knowledge extraction module and proceed to subsequent steps
• Skip Review Composition: Whether to skip the review generation module and proceed to subsequent steps

• Search Options: Open the literature search options dialog
• LLM Options: Open the large language model options dialog

• Show/Hide Options: Whether to fold the configuration options
• Run Automatic Review Generation: Start the review generation

The interface after folding the configuration options is as follows

5. Literature Search Options Configuration

1. Literature Search Options Configuration Dialog

The interface of the literature search options configuration dialog is as follows

• Add to Serp API List: Open the Serp API key (list) configuration dialog
• Add to Research Keys: Open the retrieval keywords (list) configuration dialog, i.e., using the keywords (list) for literature search
• Add to Screen Keys: Open the filtering keywords (list) configuration dialog, i.e., using the keywords (list) for filtering titles and abstracts

• StartYear: The starting year for literature retrieval
• EndYear: The ending year for literature retrieval
• Q1: Whether to search in the first-tier journals in the 2022 CAS division table for Chemistry/Chemical Engineering
• Q2&Q3: Whether to search in the second and third-tier journals in the 2022 CAS division table for Chemistry/Chemical Engineering

• Save: Save the literature search configuration

2. Serp API Key (List) Configuration Dialog

The interface of the Serp API key (list) configuration dialog is as follows

Enter the Serp API key obtained in the previous steps, one at a time, without quotes, then click OK.

Clicking OK directly or clicking Cancel will not modify the Serp API key configuration.

Enter !!!~~~!!! to clear the existing Serp API key configuration, without quotes.

The operation of the retrieval keywords (list) configuration dialog and the filtering keywords (list) configuration dialog is completely consistent with the Serp API key (list) configuration dialog.

3. Click save to close the literature search options configuration dialog.

The program interface after clicking save is as follows

Apart from the Serp API key (list), the content of other options will be printed on the interface.

If the configuration is incorrect, reopen the literature search options configuration dialog for configuration.

Directly closing the literature search options configuration dialog will not save the literature search configuration options.

6. Large Language Model Options Configuration

1. Large Language Model Options Configuration Dialog

The interface of the large language model options configuration dialog is as follows

• Add to Claude Api Key List: Open the Claude API key (list) configuration dialog
• Add to OpenAI-compatible API Url List: Open the OpenAI format API URL (list) configuration dialog
• Add to OpenAI-compatible API Key List: Open the OpenAI format API key (list) configuration dialog
• Check LLM Response: Check if the configured large language models are accessible

2. Close the large language model options configuration dialog.

The program interface after clicking save is as follows

The operation of the above configuration dialog is completely consistent with the Serp API key (list) configuration dialog.

Note that the OpenAI format API URLs and keys need to correspond one-to-one.

Supports adding multiple types of keys and can access in multiple processes to accelerate.

The configuration dialog is saved directly after clicking confirm.

After closing the large language model options configuration dialog, the number of configured large language models will be printed on the interface.

It is recommended to click Check LLM Response after configuration to check if the configured large language models are accessible.

An example of a successful test is as follows

An example of a failed test is as follows

Check the reason for the failure based on the return result and reconfigure.

Large language models that fail the test will not be applied in the review generation process.

7. Run the Review Generation Process

Refer to 4. Interface Button Introduction, select the modules to run.

Initially, it is recommended not to skip any process.

If the previous steps are not completed, there will be error prompts.

The options will automatically fold during the run.

Supports breakpoint continuation, if the run is interrupted, rerun to continue.

• The prompt for an incomplete literature search process is as follows

• 

• The prompt for an incomplete topic generation process is as follows

• 

• The prompt for an incomplete knowledge extraction process is as follows

•