Research is, for all intents and purposes, one of the most taxing parts of university education that is inevitable across all major disciplines. STEM and social science majors mostly need it for publishing journal papers, while business majors need it for competitions and academic term papers. Reading through hundreds of articles for hours – just to find that one line that perfectly aligns with the micro-niche one is working on – can get frustrating. With the advent of AI like ChatGPT or Bing, one would expect the tedious process to be alleviated. However, it was not long until these self-learning models started making up responses, providing references that did not contain anything along the lines of those responses. Yet there are websites and AI tools to make things easier, like:
Shadow Libraries
A major source of frustration is when the Google search returns an accurate result only for it to be behind a paywall. There are shadow libraries to address this pain point of students who need a citation for non-commercial purposes. These are huge online databases of content that is normally inaccessible. LibGen (Library Genesis) is one such free site that allows its users to search by title author, publisher, ISBN and other handy filters. Another free popular choice is Sci-Hub, in collaboration with LibGen. Fueled by the spirit of communism, it is the self-declared “most controversial project in modern science”. If an international paper has a DOI number, it will most likely be on Sci-Hub. The ethical implication of using shadow libraries, however, depends on the rationale of the user.
AI Chatbots
Popular general chatbots may thrive in creativity but languish in research. To address this limitation, as well as a few common research pain points, there are AI sites and extensions.
Consensus.app
Consensus answers a research question like ChatGPT does but meticulously. Its ‘consensus metre’ feature shows how many papers agree to an answer and how many don’t. It also provides exact citations from these papers, what kind of sampling was done, which one is the most cited, and allows the user to download a .csv spreadsheet file of the summaries. In the free version, there is a 20-credit limit on consensus metres and snapshots.
Perplexity.ai
Perplexity is an AI chatbot that asks the users more specific questions based on their question to provide as relevant information as possible. It summarises the answer based on multiple sources that it provides. There is also the ‘focus’ feature that narrows the search to academic papers, WolframAlpha and even YouTube videos or Reddit threads. The paid version lets the user ask questions based on any image or video they upload.
SciSpace
SciSpace lets the user ask questions from the PDF they upload. The free version allows up to 10 PDFs. Its literature review feature returns results that can be filtered to publication type (journal, paper, patent) and specific journals/conferences. It provides these features as a Google Chrome extension as well, which can analyse and summarise the webpage it is activated on.
One link to rule them all
Wouldn’t it be nice if there were a site or tool that can walk a student researcher through the long and tedious process of writing research-based assignments or papers? The GitHub project: phd-resources.github.io does just that. It compiles all the necessary sites, tools, and software to write a paper. That includes searching, generating, storing citations and bibliography. Resources include archives and repositories specific to physics, biomedicine, coding, and even HTML5 web page articles.
Research and literature reviews can be frustrating, let alone the citing and referencing that follow. But utilising these tools to the fullest can automate the processes or cut time off of going through stacks of irrelevant papers.