
Research workspace from Fudan...
Split a fuzzy topic into checks you can verify, then decide which papers deserve a place in the library. Deep Search breaks the question into keywords and conditions and marks each paper as a full or partial match, which keeps out results that only look related. Once a paper is saved, nested folders can follow a project or a theme. The text, highlights, and notes stay together, and the notes can be searched again. A watcher looks for new papers in the areas you set, and what you save or ignore feeds the next recommendations. A free account covers most of search and Q&A. Newer survey features are outside it. It fits graduate students and lab members who screen papers every day and need citations that point back to the text.
WisPaper is a research workspace from Fudan University's natural language processing lab, for finding papers, storing them, and tracking new ones. Deep Search splits a complex question into keyword sets and checks, then marks each paper as a full or partial match instead of only matching keywords. Papers you find can go into a library, be filed in nested folders, and be read online with highlights and notes that are themselves searchable. After you set research areas, the background watch looks for new papers, and what you save or ignore shapes later recommendations. A free account can use most of search and Q&A after signing in. Newer survey features are not on the free account. TrueCite on the site accepts a bibliography file and checks it against scholarly databases to flag entries that may be invented. Answers are supposed to stay anchored to passages in papers so you can check them.
Difficulty: Intermediate
Check papers against conditions
Deep Search splits a complex question into keywords and checks, then marks each paper as a full match or a partial match.
Library and notes
Papers can sit in nested folders. You can read the text online, highlight it, and take notes, and those notes are searchable.
A watch for new papers
After you set research areas and filters, a background watch looks for new papers, and saves or ignores adjust later recommendations.
Check a bibliography file
TrueCite accepts a bibliography file and checks it against scholarly databases to flag entries that may be invented and do not match a real paper.
A student screens a set of papers
A graduate student gives the topic to Deep Search and saves only the full-match papers into the library before reading.
A lab watches for new papers
Lab members set an area and filters, let the watcher push new papers, and file them into an existing folder in one step.
Check citations before submission
An author gives the bibliography file to TrueCite and checks which entries may be invented when set against the databases.
Fudan's write-up is specific. Deep Search first understands the question, splits it into keyword sets and checks, then searches with those combinations. For each result it checks the conditions one by one and marks the paper as a full or partial match. Ordinary search is closer to matching the words directly. The product page calls this intent checking: it asks whether the paper really sits inside your research logic, not only whether the same words appear.
You can save a paper into the library in one step. The library supports nested folders by project, theme, or time, and it keeps the title, authors, abstract, date, and citation details. You can read the text online, highlight it, and write notes, and the notes are searchable. A watcher looks for new papers in the areas you set, and saves or ignores affect recommendations. If a new question comes up while you read, you can start Deep Search again from the reading view.
Real reviews and feedback from users