Research services · data support for papers and projects
Before you chase the data, find out if it can be obtained
Availability comes first: we run real searches across official, institutional and well-documented sources, and report what was found, what comes close and what is missing — each item with its source.
Individual researchers · students
Data for your thesis or paper
Describe your research question in a sentence. Xiaoju helps clarify the need, starts with a free Basic Search, and then lets you decide whether to generate a Complete Search Report or request Human Deep Verification or Custom Dataset work.
Research groups · institutions
Custom data for team projects
Panel stitching, unified definitions, structured preparation and documented delivery — feasibility assessed first, delivery only once confirmed doable. Multi-source panels need aligned definitions; topic data needs cleaning and integration. Set out your criteria and the search assistant verifies matches and gaps against each required item.
We provide support for finding, assessing and organizing data; research design and writing remain the researcher's own work.
RESEARCH CAPABILITIES
Research-focused data support
Around research projects, paper reproduction and data applications, we provide research data support from availability assessment through to custom delivery.
01
Panel data preparation
Multi-year, multi-source panel data with consistent field definitions and documented gaps, supporting reproduction and secondary analysis.
Multi-year panelsConsistent definitionsReproducible
02
Literature & corpus organization
Cleaning and structuring of literature, text and multimodal corpora for research analysis and model use.
Literature corpusStructuredOrganized
03
Reproducibility material preparation
Reproducible data and processing notes, with sources and field definitions attached, so results are easy to check and cite. Research design and paper writing remain the researcher's own work.
Reproduction packageSource notesField definitions
04
Availability & compliance assessment
Before a project starts, we assess whether public data can be obtained and where the license and compliance boundaries lie, helping avoid wasted effort.
AvailabilityCompliance boundariesEvidence
05
Multi-source data integration
Alignment and integration of data across sources and across years, forming a research dataset ready for direct analysis.
Multi-source alignmentAcross yearsResearch dataset
06
Custom research datasets
We combine data tables, samples and documentation around your research goals, providing a verifiable delivery with support for later additions.
Custom deliverySample notesLater additions
DATASET SEARCH
Research edition · dataset search
Set out the data your research needs and the conditions it must meet; we search official, institutional and well-documented public sources and verify matches and gaps against each required item.
Set out your conditions, start the search
- Matches and gaps are assessed against each required item — not handed over as a pile of links.
- Real searches covering official, institutional and well-documented public sources.
- For matches, a structured dataset reference card plus source link; where there is no direct match, we still present search directions and near-equivalent sources.
The search is run by our agent; we provide source links, the search process and the results, and do not deliver raw data files. Need a usable dataset delivered? The team can build one to order (quoted per request). The conversation is only for discussing research data needs and is not a formal commitment.
How it works · how it's billed
- Basic Search — Free: quickly assess whether relevant public data exists, with main candidates, source leads, preliminary matches and clear gaps. It is automated and does not mean files were downloaded or their contents and licensing were checked.
- Complete Search Report — RMB 10 per run / 1 credit: expands specialist terms, Chinese and English names, abbreviations, aliases, search paths and public sources, with a fuller candidate list, per-condition search-stage assessments, main gaps and next steps.
- Human Deep Verification — Quoted separately: checks actual access routes, file structures, fields, samples, licensing conditions and usage barriers for priority candidates.
- Custom Dataset — Quoted after assessment: covers collection, cleaning, integration, labelling, quality control and formal delivery when public data cannot meet the required volume, fields, years, regions, formats, labels or quality.
- Viewing an availability assessment report requires signing in (it contains dataset names, source links and per-item verdicts); results are not shown to signed-out visitors.
- Basic Search works from the need you describe, searches accessible public platforms, academic and industry materials and institutional sites, and makes search-stage assessments against each key requirement.
- The deliverable of a standard search is an availability assessment report: candidate sources, match verdicts, gap notes and next-step suggestions.
- As an information-retrieval service, we cannot promise a perfectly matching dataset every time. We widen the search around your requirements and document the results and reasoning in the report.
- Complete Search Reports are billed per run, not by whether an exact match is found. The service is complete once a readable report is generated; a no-match is not refundable. No credit is charged when a system issue prevents a readable report, and an abnormal charge is refunded or compensated through the existing process.
- If a system or platform fault prevents the report from being generated or viewed, or the report clearly falls short of the service description on review, we refund or compensate the credits.
- If nothing fits closely enough, leave a note and the team will assess a custom dataset (collected and prepared around your required conditions), quoted per request.
- Credits do not expire.
Curated open datasets · structured reference cards
Open the dataset library →
Penn World Table, bilateral trade, nighttime lights, global land cover… high-value open datasets organized into structured cards covering contents, fields, use cases, access paths and license notes. Not among them? Set out your conditions — we assess availability first, then run a real search.
DATA-FINDING GUIDES
How to find public data
The common questions of finding public data, organized into guides: where to look, how to choose a platform, whether a license allows commercial use, and how to look up official data.
Where to find research data
By discipline, a list of legitimate public data sources spanning official platforms, university subscriptions and international repositories.
Comparing open dataset platforms
How to choose among common open data platforms by data type, including license characteristics and citation methods.
Where to find ML training datasets
Practice competitions, training corpora, classic benchmarks and large-scale corpora, platform by platform.
What panel data is, and where to find it
The difference between panel, cross-sectional and time-series data, with common public panel sources for economics and social science.
Can open datasets be used commercially
How to read licenses such as CC and ODbL — commercial use and attribution requirements in one clear table.
How to look up official statistics
How to access and export from the National Data platform and the China Statistical Yearbook, and what to watch out for.
Four steps to a result
01
Describe your need
One sentence to start, plus the conditions that must be met and the preferences worth noting.
02
Real search
We search and screen across well-documented data platforms; matches are never fabricated.
03
Assessment report
What was found, what comes close and what is missing — each item with its source.
04
Obtain under license
You obtain the data under each source's license; optional custom preparation and documented delivery.