CMU-Qatar Library

In a Times Higher Education article by Da Silva and El-Ayoubi (2023), the authors emphasize the potential of GenAI in academic research while also highlighting the need for human expertise, knowledge, and critical thinking in evaluating and refining what AI has generated. The article discusses the impact of GenAI models on higher education and academic research. They propose three critical steps for integrating these tools into research practices while acknowledging their limitations:

a) Refining Research Questions

• AI can help brainstorm ideas and identify research gaps.
• It can assist in literature reviews and data exploration.
• Human expertise remains crucial for evaluating AI-generated suggestions.

b) Enhancing Research Efficiency

• AI can summarize research methods and rank their effectiveness.
• It can analyze large volumes of text and generate reports.
• AI can aid in data analysis and visualization.

c) Supporting Research Validation

• AI can suggest validation frameworks and methods.
• It can help identify potential issues in data or methodology.
• AI can provide feedback on research findings and writing.

The authors emphasize that while GenAI can be a powerful assistant, it lacks true understanding and should not replace human expertise. They suggest treating AI as an assistant and not as a substitute for the disciplinary knowledge expert. They also stress the need for human review and validation to ensure accuracy, interpretability, and relevance of the generated outputs.

Dimensions

Dimensions, the world's largest collection of linked research data, has launched an AI-driven summarization feature to enhance discovery of publications, grants, patents, and clinical trials. This new capability allows to quickly access concise, AI-generated summaries for each record in the search results.

Keenious

Keenious is a resource recommender tool designed to aid in the identification and discovery of scholarly research. It analyzes writing from text documents and PDFs to recommend the most relevant articles and topics to explore using AI combined with conventional search algorithms. It seamlessly integrates with library resources to minimize paywalls, and is freely available to all Carnegie Mellon affiliates.

Scite

Scite is an AI-powered research platform that enhances the evaluation of scholarly articles by analyzing citation context. Its key feature, Smart Citations, categorizes citations as supporting, mentioning, or contrasting the cited work, allowing researchers to understand how a paper has been referenced. The Scite Assistant enables users to ask research questions in plain language and receive answers directly from over 32 million research articles.

Scopus

Scopus has introduced a new AI feature, Scopus AI, designed to enhance the research experience by allowing users to search using natural language queries. This innovative tool provides concise summaries, expanded insights, and concept maps based on the extensive Scopus database, which includes over 27,000 academic journals and 1.8 billion citations. Scopus AI highlights foundational papers and experts in specific fields and suggests follow-up questions for deeper exploration. The feature is currently in a trial phase, available to all users, including those using the free version of Scopus.

Notice: You can access these databases through CMU Libraries Databases A-Z. Make sure you use your Andrew ID and password to create your account.

Created using Claude Sonnet 3.5

Notice: Some of these tools have faced criticism for their use of copyrighted content without prominent attribution and have limitations in fully grasping context compared to human experts. The generated AI summaries can sometimes lack full contextual accuracy compared to human interpretation. Therefore, you are required to approach the use of these AI tools with a critical mindset, ensuring that you verify the information presented and cross-reference it with original sources. It is essential to maintain a rigorous standard of academic integrity by properly attributing any content derived from these tools and acknowledging the limitations inherent in AI-generated outputs.