Refer to the video for a visual guide
Setting up Data Extraction in EasySLR is simple and straightforward. You can choose to set up data extraction while setting up the project, or later after you finish Title abstract and Full text screening.
We have provided a few preset templates to make the task easier for you. You can easily customise these templates to fit your project's specific needs.
To access and customise available templates, click on "Add New Sheet." From there, you can select a template that suits your needs.
Additionally, you can add custom data fields by clicking on the "Add data extraction item" option. Define the field name and provide a brief description of the specific information you want the AI to extract.
This flexibility allows you to tailor the extraction process to your specific project requirements with ease.
Reorder function allows you to adjust the sequence of the fields. This feature allows you to streamline and customise the order of fields for optimal management and editing flexibility.
We offer additional options, such as adding multiple sheets for different outcomes and including specific data fields. Each sheet can be named based on the type of data being extracted within it.
You also have the option to upload your own data extraction template. For guidance, we provide a sample template as a reference to help you create your own.
After updating the data extraction template, activate AI functionality to enable automated extraction of all available information.
To validate AI extraction, go to the “Data Extraction” section under Review stages.
Data Extraction is divided into three sections to categorise studies based on their extraction stage. Initially, all articles will be categorised as "Articles to Extract," signifying that the extraction process has not begun.
Click on “Start Extracting” to view the data extracted by the AI. Check the extracted data for accuracy and completeness.
You'll find the extracted data alongside each field in the right-hand side pane. This panel also provides links to the specific sources (one or multiple) within the PDF where the AI sourced that information.
This feature facilitates quicker quality control (QC) and validation, allowing you to easily verify the accuracy of the extracted data.
Additionally, you can edit the extracted data, to make necessary changes and updates as required. The linked sources associated with the extracted data are also editable allowing you to add or delete sources as needed.
To add sources, click 'Edit' next to the existing sources for the field. Select the relevant data from the PDF, click '+', then 'Done'. The new source will be linked.
Once you've completed the extraction and made necessary updates, select ‘Extraction done’ available at the bottom of the screen and click ‘Update’. This action moves the article to the "Extraction Done" section, where it will be available for another reviewer to begin the Quality Control (QC) process.
Note: Using the "Update & Next" button allows reviewers to proceed to the next article while keeping the current article in its respective extraction stage.
In the “QC Done” section, the reviewer will have access to the finalised extracted data, along with detailed information about the individuals who contributed to the data extraction at various stages. This ensures transparency and accountability for each step of the process.
Once the QC process is completed, the reviewer selects "QC Done" and clicks "Update." This action moves the article to the "QC Done" section, indicating that the work on that study is complete.
If rework is needed, select 'Pending' and 'Update' to return the study for extraction.
Once the extraction process is complete, you can easily export all the data into an Excel file for further analysis and reporting.
Comments
0 comments
Please sign in to leave a comment.