Literature Search and Filter

A Python tool to streamline literature screening using automated filtering by page limit, language, duplicates and boolean-logic-based keyword search.

Supported Databases:

• ACM Digital Library
• IEEE Xplore
• ScienceDirect

Getting Started

Export your database search results as BibTex files and place them in a suitable directory.

Supported Python version >= 3.10.0. Install the dependencies by running the following command.

pip install requirements.txt

Usage

Run python main.py with the below arguments.

Argument	Value	Required
--file_path	Path to your BibTex Directory	✓
--mode	"acm", "ieee", "scdrt"	✓
--page_limit	Preferred Minimum Page Count	✓
--excel_path	Path to Output Excel File	✗
--search_keywords	None	✗

On completion, an Excel file will be created containing the information (Author, Title, DOI URL) on literature that successfully passed the screening process. If --excel_path is specified, the selected Excel workbook will be appended with the results of the screening process; this may be useful if using multiple databases in a single literature review. Otherwise, a new file will be created.

Keyword Boolean Logic

If using the keyword search functionality, the tool will prompt you to do the following:

• Enter required keywords separated by a comma (,) to run OR operator.
• Add keywords on a newline to run AND operator or type "X" to finish.
• To search for an exact match of a keyword, enclose it within quotes (" ")

For example, the Boolean logic ("Technology" OR "Computer" OR "Computers" OR "Computing") AND ("Program" OR "Programs" OR "Programming" OR "Coding") would be entered as:

"Technology", Comput
Program, "Coding"
X

Advanced Usage

The tool, by default, filters for publications in English whose Title and Abstract satisfy the defined conditions. Slight modifications to the source code can allow for a more fine-grained search strategy.

Language Criteria:

condition = lambda entry: page_key in entry and (int) (entry[page_key]) >= PAGE_LIMIT and detect(entry[title_key]) == "en"

Publication Data Extraction:

extract_text = lambda entry: " ".join([entry.get("abstract", ""), entry.get("title", ""), entry.get("keywords", "")])