Frequently Asked Questions

Artemis • BiblioSelect is a platform for article screening in systematic, integrative, and scoping reviews. You create an account, import a CSV with the results of your literature search, and classify each item as useful or not useful — manually or with the help of AI (optional, paid via credits). At the end, you export the selected items and, on the Pro plan, generate a report with a PRISMA flow diagram ready to paste into your paper, thesis, or dissertation.

Yes. Signing up is free — just email and password, no credit card. Your account is what stores your projects, files, and screening decisions, so you can access them from any device.

Two plans, no mandatory trial period:

• Free — R$0: up to 3 projects, each with up to 200 articles. Full access to manual screening and to AI (if you buy credits).
• Pro — R$29/month: up to 10 projects, each with up to 1000 articles, plus access to the Report tab (PRISMA flow diagram, automatic charts, AI×human agreement, and methods paragraph).

AI credits are billed separately, on any plan — see the credits question below.

Open Account (top menu) and click Subscribe to Pro — you'll be redirected to a secure Stripe checkout. To cancel or change your payment method, go back to Account and click Manage subscription, which opens the Stripe billing portal. You can cancel anytime; Pro access continues until the end of the period you already paid for.

Each credit pays for the AI review of 1 article. Credits are sold in one-time packs (not a subscription), work on any plan (including Free) and in any project, and never expire:

• 100 credits — R$15
• 250 credits — R$30
• 500 credits — R$50
• 1000 credits — R$90 (best price per credit)

Purchase from the Account tab, under AI Auto-Review Credits, also via Stripe checkout.

Use the CSV exported from SCOPUS. The app reads the standard headers from that database. The most useful fields are:

• Title — article title
• Abstract — abstract
• DOI — identifier (optional)
• Cited by — citation count (optional, numeric)

For now we only support the SCOPUS format. In future versions, we plan to add a field mapping step for other databases (e.g., Web of Science, PubMed), while preserving original columns in the export.

1. Open Screening (after creating/opening a project).
2. Use the Upload CSV button at the top.
3. The file is processed in your browser; you will see the list in the left-hand column.

Projects organize your reviews into separate workspaces. Each project keeps its own CSV and screening state (useful items, hidden items, AI criteria, query, and sorting) without affecting the others. On the Free plan you can have up to 3 projects with up to 200 articles each; on Pro, up to 10 projects with up to 1000 articles each.

In the left column, you see the filtered and sorted list. When you select an item, the center panel shows the title, abstract, and actions:

• Add to useful — moves the item to the selected list (right-hand column).
• Remove from list — hides the current item from screening (this can be reversed later by clearing the state).
• Copy reference — copies title/DOI/citations to the clipboard.

The search bar (top toolbar) filters by title and abstract. Sorting can be by citations (↑/↓) or title (A–Z/Z–A).

It's a way to speed up screening: the AI classifies pending articles as useful or not useful, based on criteria you define. Steps:

1. Set your inclusion/exclusion criteria (a list of statements, reorderable) via the criteria button on the Screening page.
2. Click Start AI. Each still-pending article consumes 1 credit and gets classified.
3. For each item, the AI shows a suggested decision, a confidence level, and a short reasoning.

The AI's decision stays fully reviewable: you can reclassify any article manually at any time, and the AI's reasoning is kept even after you override the decision. With no credits available, the button takes you straight to buying credits.

The Results page becomes available once there are no remaining items pending classification in Screening. It has six tabs:

• Overview — consolidated screening numbers (total screened, included, excluded).
• Sources — distribution of included articles by source (journal/venue).
• Authors — most frequent authors among included items.
• Documents — browsable list of included articles.
• Report — PRISMA flow diagram and ready-made charts (exclusive to the Pro plan, see the next question).
• Account — shortcut to manage your plan and credits.

Yes, the Report tab is exclusive to the Pro plan. From the data already screened in your project, it generates:

• Automatic PRISMA flow diagram — from total screened to included and excluded, in the format systematic reviews and top-tier journals typically require.
• Automatic charts — included by thematic axis and excluded by reason, ready to paste into your results section.
• AI × human agreement — agreement rate between the AI's suggestion and the final reviewed decision (useful as evidence of methodological rigor, if you used Auto-Review).
• Methods paragraph — ready-to-paste text for your paper, thesis, or dissertation.

Everything is generated in seconds from the screening you've already done — nothing needs to be reprocessed.

In the right-hand column of the Screening page (Selected), click Download CSV. The file will contain only the original rows marked as useful, preserving the headers from your input file.

Yes. Unlike the old version (100% local), BiblioSelect is now a multi-user platform: your CSV, screening decisions, and project state are stored in a cloud database, tied to your account. That's what lets you access the same project from any device.

When you use AI Auto-Review, the title and abstract of each classified article are sent to our AI provider only at the moment of classification — this only happens if you start Auto-Review, never during manual screening.

RESET clears the current project's data: uploaded file, labels (useful/hidden), query, sorting, and current selection — without deleting the project itself. This action is irreversible. Use with caution.

From the home page or inside the project, use the Delete project option. This removes the project, its uploaded file, and all associated screening decisions — irreversibly. Unlike RESET, this removes the project from your list and frees up a slot under your plan's limit.

• That the file is actually a .csv.
• That the first line contains a header row.
• That there is at least one column with a title and another with an abstract (or equivalents) — this improves search and display.
• That the separator is correct (comma). Some exporters use semicolons; in that case, convert or re-export as a standard CSV.

Describe the problem (including whether it's about screening, billing/Stripe, or account access) and — if possible — include a snippet of the CSV involved (without sensitive data) and any browser console error messages. This also applies to password-change requests.