AI supported Search Builder

Problem Addressed:
Developing and executing literature search & review strategies is time-intensive, taking up to over 100 hours.

Solution:

• An AI-driven tool to automatically generate search strategies based on user inputs such as key concepts, keywords, or a foundational article.
• Machine learning algorithms analyze user-selected key articles to suggest related terms, synonyms, and database-specific thesaurus terms.
• A user-friendly interface that allows users to refine and adjust search strategies without requiring deep technical expertise.

Smart Search String Generator

Problem Addressed:
Search strings are complex, differ across databases, and are difficult for young researchers to master.

Solution:

• A syntax-agnostic engine capable of generating search strings tailored to each database’s requirements, including AND/OR logic, syntax formatting, and thesaurus integration.
• Built-in guides and visual aids for constructing and understanding search strings.
• An autocomplete feature for terms, operators, and syntax, reducing manual work and errors.

Cross-Database Translation Tool

Problem Addressed:
Translating search strings between databases (Embase, PubMed, Ovid) is challenging due to unique thesauruses and search engines.

Solution:

• A translation engine that maps terms and syntax between databases using their specific thesaurus.
• AI algorithms to suggest equivalent terms, alternative phrasing, and adjustments for each database’s unique structure.
• Integration with major databases to ensure real-time thesaurus updates and compatibility.

Semi-Automated Structured Search Workflow

Problem Addressed:
Regular processes involve extensive manual work and interpretation. Making searches, running de-duplication, consolidating results, reviewing etc. It can be tedious, person depending, error sensitive.

Solution:

• A guided workflow system that automates repetitive tasks like de-duplications, translations and report generation.
• Decision-support tools to assist with steps requiring interpretation, such as suggesting search terms based on a key articles.
• Progress tracking to monitor the completion of each step, ensuring users maintain a structured approach without manual oversight.

Training and Support Features

Problem Addressed:
Young researchers face a steep learning curve in building search strings.

Solution:

• Interactive tutorials and step-by-step guides integrated into the software.
• Real-time feedback on search string accuracy and suggestions for improvement.
• A “sandbox” environment for testing and refining search strategies without risk.

---