Opening Research Compass
See what is already published on your topic before you choose it
Type your topic in plain words. Thesisroom counts every matching paper in the OpenAlex research database and shows where the field is crowded and where it is thin.
- Three searches a day are free, and no card is needed
- A result, usually in under a minute
- Nothing is written for you
Free to start, no card needed. Then $9 a month billed yearly, or $12 month to month.
What are you looking for
Choosing your subject? Type it and see what has been published on it, how the papers group, and where to start reading.
Explore a topic when you are choosing your subject. Check a claim when you know what you want to argue.
It counts and lists published papers from OpenAlex. It does not tell you what to study.
- DataCite
- OpenAIRE
- HAL
It gave you papers that do not exist
You typed your topic into ChatGPT or Claude or Gemini and it gave you a reading list that looked perfect. Real-sounding titles, real-sounding authors, real-sounding journals. Then you searched for the first one and it was not there. The DOI went nowhere. The author never wrote that paper. The journal exists but that article does not.
That is not a bug. That is how those tools work. They write sentences that sound like citations. They do not check whether the paper is real, because they have no connection to the places that hold real papers.
What ChatGPT does
- Generates text that reads like a citation
- Invents titles, authors and DOIs from patterns in its training
- Cannot check whether a paper exists
- Cannot tell you if a source was retracted
- Gives you a different list every time you ask
What Research Compass does
- Searches OpenAlex for a topic, and Semantic Scholar for a claim
- Returns only papers that are already in the record
- Every paper is a real record you can look up yourself
- Shows how often each paper has been cited, from the record itself
- Shows the same record every time, because it is the record
One writes what sounds right. The other looks up what is there. That is the whole difference.
Everything below can be checked against the record, down to the questions
- The papers that match your topic, sorted into groups, with the count covering the whole record
- Up to 150 papers on your words, each one a real entry you can click and read, with the 15 to read first picked out
- The authors who appear most, with their institution, taken from the papers themselves
- How many papers were published each year, so you can see whether a field is growing or has gone quiet
- Up to three questions your topic raises, each one naming a specific paper from your results
- A one-page summary you can print and take to a supervision meeting
- Your search saved permanently so you never run it twice
- The same search runs again every week while you use it and marks what appeared since, with an email of the new papers if you ask for one
If the record holds nothing near your words, it says so. Nothing here invents a result to look useful.
Someone typed "green finance and sustainability"
This is what came back. Eight groups with paper counts across the whole record, the ten most-cited papers, the five newest, the key authors, and three questions the topic raises. Every paper links to the source. Every question names the paper that raised it.
Free. Nothing is written for you.
green finance and sustainability
Sustainable Solutions for Green Financing and Investment in Renewable Energy Projects
Research on the impact of green finance on the upgrading of China's regional industrial structure from the perspective of sustainable development
'Green financing and sustainability reporting among SMEs' links pro-environmental behavior to digitization. In your own context, is the SME angle a direction your research centers on, or does it sit outside what you are trying to understand?
The first month of reading decides which papers your argument stands on
You pick a direction from what your supervisor mentioned and what the first page of Google Scholar returned. Three weeks later you find the field moved, or that the question has two hundred answers already. That is a month of reading that pointed nowhere, and most of it was findable before you started.
- 01Type what you are thinking aboutPlain words, not keywords. A sentence about your interest is enough.
- 02Thesisroom searches published research databasesReal databases holding real papers. The papers are never invented or guessed.
- 03You get a map of the field, not a suggestionPapers sorted into groups, the key authors, a reading list, and the questions your topic raises.
Free. Nothing is written for you.
Three people who ran their own research through Thesisroom
What happens to your search
Your search is saved in your account. It is sent to OpenAlex to find the papers and, once you are signed in, to the language model that reads it; it is never used to train any model, and you can delete it permanently at any time.
What this will not do
Thesisroom does not suggest a topic, does not recommend a methodology, and does not say whether a direction is promising. It shows what the record holds and asks up to three questions. The answers are yours.
Free to search. Twelve dollars to keep going
Questions researchers ask
Do I need an account to search
How is this different from asking ChatGPT about my topic
How do I find out where to start on a research topic
How do I see what has already been studied
How do I find out who is publishing in my field
How do I know if a direction has been exhausted
What if my idea is too vague
- Find Sources takes one claim and returns papers that speak to it, each a real entry you can look up
- Blind Spots reads what you have written and names the work the record ranks highly that your sources do not contain
- Scope Check reads a proposal and quotes the two passages where what it asks and what it plans sit apart
- Reading Queue orders what to read next, starting with the claims the record returned least for
- Reference Watch runs this same weekly check over the sources you have cited, and tells you when one changes
See what the record holds before you commit to a direction
What you get back
Type what you are thinking about and Thesisroom searches OpenAlex, a published research database, for papers whose title or summary matches your words. It sorts what comes back into groups, using the labels the databases already assign, and shows how many papers sit in each group across the whole record rather than in a sample.
Beside that you see how many papers were published each year, the authors appearing most often across the results, the ten most-cited papers on your words and the five newest. Every paper is a real entry with an identifier you can look up, and nothing on the page is written or summarised for you.
Why it shows counts instead of advice
A group holding nine papers beside one holding two hundred tells you something, and what it tells you depends on things no database knows: your supervisor, your programme, the methods you can use. Thesisroom shows the counts and stops there. Whether a quiet corner is a question nobody has asked or one that turned out to be a dead end is yours to work out, and it is the question the tool asks you rather than answers.
The questions at the bottom of your results work the same way. Each one names a specific paper from your own results as the thing that raised it, and one of them presses on the thinnest part of what came back. A question that cannot name a paper from your results is dropped before you see it.
Which of your papers cite each other
Research Compass draws the papers found for your search as a citation network. Each paper is a circle, coloured by its group and sized by how many of the other papers found cite it. Each solid line is one paper citing another, and each dashed line two papers citing three or more of the same sources. No paper is added.
A map with no lines is a result in itself: the papers closest to your words do not cite one another, which can mean the groups sit further apart than the shared keywords suggest. The connections come from the reference lists the papers themselves publish, so nothing is estimated and no paper appears that the search did not find.
Checking a claim rather than a topic
Research Compass also checks a claim. Type one sentence, such as whether a style of leadership improves team performance in remote work, and it lists the papers whose text speaks to that sentence, in Semantic Scholar's own relevance order, each read for language about both sides of the claim.
Each paper's summary is scanned for result language in sentences that name both sides of the claim, and a badge says whether it carries supporting or questioning wording, or neither. The badge is a count of words, not a verdict. The sentence that produced it sits under the paper title with those words marked, so you can read the evidence for the label without opening anything. Above the list sit three counts and nothing else. Whether the evidence supports the claim is a judgement the summaries let you make, and the tool stops before making it.
Questions people ask
The things a researcher wants to know before they commit three years to a subject, and what this page can and cannot tell them about each.
- How do I find out where to start on a research topic?
- Research Compass takes the words you type and shows the published papers already near them. You read the most-cited papers and the newest ones, and pick your own starting point from real entries.
- How do I see what has already been studied on my topic?
- Research Compass sorts every matching paper into groups and prints how many sit in each, across the whole record rather than a sample. What is crowded and what is sparse is something you can see for yourself.
- How do I find out who is publishing in my field?
- Research Compass lists the authors who appear across your results, with their institution and how many of those papers they wrote. The names come from the papers themselves, never from a ranking of people.
- How do I know whether a research direction has been exhausted?
- Research Compass shows the paper count and the publishing years for each group, and stops there. A thin area can mean the question is settled or that nobody has asked it, and no database knows which.
- Can I check whether the evidence supports a specific claim?
- Research Compass has a claim check: type a claim and it lists the papers whose text speaks to it, each with the sentence from its own summary that carries supporting or questioning language, or neither. The reading is yours.
- What is the difference between Explore a topic and Check a claim?
- Research Compass has two modes. Explore a topic when you are choosing your subject. Check a claim when you know what you want to argue.
- How do I see which papers in my results cite each other?
- Research Compass draws the papers found for your search as a citation network. Each circle is a paper, coloured by its group. A solid line is one paper citing another, a dashed line two sharing three or more sources.
What happens after your search
Your search is saved to your account, so the result is there when you come back to it rather than something to run again. The same search runs again every week while you use it and shows you what appeared since, and you can ask for the new papers by email. When you have written something, the rest of Thesisroom reads it against the same published research databases.