Great writing by Scholingo when it comes to landing on the shore

What is Scholingo, and how is it excellent at what it does?

Kaoan Miaoxie Scholingo is an AI-based academic writing platform operated by Shanghai Yida Network Technology Co., Ltd.; it offers services such as the creation of draft papers, AIGC detection and rewriting, Word format checking, and text paraphrasing tools.

It is useful for helping to choose topics, outline the structure, create a preliminary draft, and check the formatting, but it cannot replace research, experiments, literature review, the author’s judgment, or the university’s checks regarding academic integrity.

Current Features Overview

AI paper writing

Start with the research topic.

• Users describe the research topic, question, or a more detailed research concept in the input field, and the system uses this information to create the paper.
• The topic should include the research subject, variables, methods, context, and objectives, and it is necessary to avoid simply entering a broad academic field name.
• High-risk topics such as medicine, law, finance, and engineering require the use of authentic literature, data, and professional review.

Word count, academic level, and language

The current settings do not include an option for a doctoral degree, nor are there any clear specifications regarding discipline-specific models, word-count errors, limits on context length, or the number of times regeneration is allowed.

Four-step writing process

1. Enter the core research question, and select the target word count, academic level, and language.
2. Let the system analyze the topic and suggest research directions, then review and revise the outline.
3. After confirming the sections, a complete paragraph and a draft will be generated; do not skip the structural review.
4. Use detection, rewriting, and formatting tools to handle expressions and layout, and then verify the factual statements and citations manually.

Research framework and outline

• The system organizes chapters, arguments, and research directions around the chosen topic, helping users to start writing from a blank page.
• Users should check whether the research questions can be answered, whether there is any repetition among the sections, and whether the methods support the conclusions.
• The concepts, samples, methods, and variables outlined must be consistent with the actual research; the model should not create a fictitious research design.

Chapters and draft of the paper

• Once the framework is confirmed, the platform can proceed to write complete paragraphs and sections, resulting in a draft with a length of 3,000–10,000 words.
• The generated content can be used to organize expressions and identify structural issues; it should not be submitted directly without being revised and verified first.
• Models may fabricate facts, experiments, data, survey subjects, and research results; the author must remove any content that has no real basis.

Citations and references

• The product page mentions the addition of citations and the adaptation of citation formats, but it does not disclose the specific databases of references, the scope of searches, or the mechanisms for verifying citations.
• The generated author, title, journal, year, volume/issue, page numbers, and identifier may be missing or inconsistent with each other.
• Each citation must be checked against the original source material and the full text to confirm that it indeed supports the corresponding argument.
• The literature review feature is currently marked as upcoming and should not be considered a standalone workflow that is already available.

AIGC detection

Short text detection

• The input field in this demonstration allows for manual entry or pasting of Chinese text, with a maximum display length of 800 characters per field.
• After initiating the detection, the user can view the AI-generated probability score provided by the platform and then decide whether to make any changes.
• Short-text results are suitable for comparing different versions of the same paragraph, and should not be used as the sole evidence of an author’s identity or academic misconduct.

Full-text detection

• The page provides a tool for full-text detection and rewriting, making it suitable for handling papers that exceed the limits of the input fields available for demonstration purposes.
• Completing the entire task will consume the character quota purchased; the specific counting method and acceptable file formats are indicated on the submission page.
• Before uploading, sensitive information such as name, student ID, supervisor, university, respondents, and unpublished research should be removed.

How to interpret the test results

• Due to differences in the models, thresholds, training data, and versions of various detection systems, the same text may yield different results.
• Manually created content can also be misjudged, and AI-generated content may receive a low likelihood score.
• The platform claims to be compatible with testing environments such as CNKI and Wanfang, but it cannot guarantee that third-party platforms will assign the same scores.
• The detection algorithms can change, and any claim stating a fixed pass rate or a value below a certain threshold should not be considered as a guarantee of acceptance.

Reduce AI rewriting

How does the function work?

• The AI rewriting function restructures sentence patterns, words, and semantic density to produce a version that is closer to natural expression.
• Users can re-evaluate the rewritten result to compare the degree of similarity, readability, and any changes in the original meaning.
• The package instructions clearly state that both the rewriting step and the verification step are counted toward the word count.

Required checks after rewriting

• Check each sentence for facts, terms, numbers, formulas, causal relationships, restrictive conditions, and the placement of citations.
• Rewriting may alter the research findings or turn cautious statements into absolute assertions; it should be in line with the actual evidence.
• Technical terms, scale names, legal provisions, and direct quotes should not be replaced arbitrarily in order to lower scores.
• Lowering the detection score does not mean increasing originality, nor can it be used to take someone else’s work and present it as one’s own.

Paper plagiarism reduction tools

• Plagiarism reduction tools are used to rewrite repetitive expressions, and they are suitable for working with text for which citations have been added and the research has been completed.
• Appropriate reduction of word count should be achieved by reorganizing the arguments and expressing them in one’s own words, rather than simply replacing words with synonyms.
• Even if the quoted content is rewritten, it may still need to be cited; the actual author or source of the material cannot be concealed by means of text simplification.
• The plagiarism rate, the likelihood of being AI-generated, and the quality of the paper are distinct indicators; a low score does not indicate that the academic content is reliable.

Word paper format checking

Supported files and outputs

• Format checking currently supports only DOCX files, with a maximum size of 20MB per file.
• The system checks the paper’s format according to specific rules, and generates two Word files: a revised version and a version with annotations.
• The verification process will not overwrite the original text that was uploaded, but users should still keep a local backup before uploading it.
• The generated file needs to be reopened in Word on the desktop or in the software specified by the school, in order to check the layout and compatibility.

Default rules and custom configurations

• The default settings are based on general national standards and those of leading universities, but the templates may vary from school to school, college to college, and from year to year.
• Users can paste their own formatting requirements, which the system will parse and apply to the inspection process.
• The rule editing field currently shows a maximum limit of 1,000 characters; it can be cleared or re-parsed.
• The page indicates that the parsed rules are saved locally and will be displayed automatically upon the next access.

Configurable format content

• Examples of rules cover abstracts, keywords, tables of contents, headings at all levels, the main text, references, appendices, and acknowledgments.
• It allows for the specification of font type, font size, bolding, centering, first-line indentation, hanging indentation, line spacing, and margins before and after paragraphs.
• Complex tables, formula numbers, headers and footers, table of contents fields, sectioning, figure captions, and special templates still require manual inspection.
• The automatically corrected version may alter the local styles; it is not sufficient to assume that the file meets the school’s requirements just because it has been successfully generated.

Format validation process

1. Obtain the official thesis template and formatting requirements for the current year from the school or college.
2. Copy the rules for headings, text, abstract, table of contents, references, and appendices as specified in the requirements.
3. Open the rule configuration, paste the content and carry out parsing; then check whether the rules resulting from the parsing are complete.
4. Keep the original DOCX backup, and upload a paper file that is no larger than 20MB.
5. Start the check, and download the corrected version as well as the annotated version of Word.
6. Verify each item against the annotations, paying special attention to tables, formulas, pagination, the table of contents, and page numbers.
7. Review it in the software and printing environment designated by the school, and then generate the final submission file.

Price packages

The page for the 100,000-character package shows both the original price of 199 yuan and the limited-time discount; the package that includes 3 papers has an original price of 117 yuan. The promotions may vary depending on time, channel, and account.

Character packages can be used for detection, suppression of AI-driven revisions, and writing assistance tools; the total amount allocated for these purposes is less than the overall quota. As for how to use the remaining quota, it should be determined in accordance with the account rules and the payment instructions related to the tasks.

It is necessary to confirm before making a purchase.

• Confirm whether you are purchasing a count of papers or a package of characters; the purposes of these two types of quotas are different.
• Check whether rewriting, the initial review, and the re-review each deduct characters, as well as whether failed tasks are returned.
• Check the number of rule parsing attempts, the number of times the final version was downloaded, the file size, and the expiration date.
• The page provides a link to the purchase agreement, but in the public environment the details regarding refunds, invoices, and dispute resolution are not fully displayed.
• It is not yet clear whether a refund is possible once digital content has been generated; therefore, it is necessary to save the order page and the relevant terms before making the payment.

Platform and language support

Which users are it suitable for

• Undergraduate and graduate students who need to quickly develop an outline and a first draft based on their research questions.
• The research has been completed; authors of papers wishing to have their structure, language, and Word formatting checked are welcome.
• Users who need to compare the AIGC likelihood of a Chinese paragraph with its rewritten version.
• Graduates are required to produce revised and annotated versions in Word in accordance with the school’s rules.
• Non-academic users who write work reports or general texts and wish to improve their expression.

Product advantages

• The paper follows a step-by-step approach, covering the theme, direction, outline, and chapters, which facilitates making adjustments to the structure at any stage.
• The word count, academic level, and language options are clearly defined, making it suitable for drafts of various lengths.
• Detection, rewriting, and formatting checks are all handled under the same account, reducing the need to switch between files.
• The formatting tool outputs both the revised version and the annotated version, which helps to identify the locations of automatic changes.
• Rules can be pasted and parsed by users, offering greater flexibility than using fixed school templates.

Limits of capability and academic risks

• The generated text may contain false citations, incorrect facts, fabricated data, and unfounded research conclusions.
• The bachelor’s or master’s options only serve to create style settings; they do not demonstrate that the quality required for a thesis has been achieved.
• AIGC detection is not a method for identifying the author; it can lead to false positives, false negatives, and differences in scores across different platforms.
• \"Reducing AI\" can only alter the characteristics of text; it cannot turn unresearched content into genuine academic work.
• The literature review has not yet been published, so it is not possible to rely on the platform to carry out automated system searches, as well as the selection of relevant studies and the synthesis of evidence.
• Format checks cannot replace the final review conducted by the institution, especially with regard to formulas, charts, tables of contents, page formatting, and references.
• Different schools have varying rules regarding AI assistance, disclosure, attribution, and acceptable uses; it is necessary to consult one’s own school’s policies before using such technologies.

Privacy and data security

• The platform claims to use encrypted transmission to protect documents, but statements such as \"no data will ever be leaked\" should not be regarded as guarantees of zero risk.
• Papers may contain unpublished results, subject data, corporate information, personal identities, and intellectual property.
• Before uploading, the name, student ID, supervisor’s details, contact information, signature page, original data, and any information that could identify the participants must be removed.
• The public page provides links to the service agreement and privacy policy, but the retention periods, model training processes, third-party processors involved, and deletion timelines have not yet been fully specified.
• Confidential research projects, materials prior to patent applications, data subject to ethical restrictions, and confidential documents of partners should not be uploaded when the terms are unclear.
• The formatting rules are stored locally in the browser; on public computers, it is necessary to clear these records and log out of the account.

Copyright and academic integrity

• The author is responsible for the research design, data, arguments, citations, and the final submission.
• When citing others’ opinions, data, images, and scales, it is necessary to provide proper attribution even if the content has been rewritten.
• Do not use plagiarism reduction or AI mitigation tools to conceal acts of plagiarism, having work written on one’s behalf, fabricating data, or violating course requirements.
• The details regarding copyright, commercial rights, and exclusivity of the generated content have not yet been made public; they need to be confirmed before the content is published or used for commercial purposes.
• When submitting a paper, it is also necessary to comply with the journal’s rules regarding generative AI, author qualifications, disclosure requirements, and data availability.

API, SDK, and open-source status

• Currently, there are no confirmed public API documents for developers, options for key acquisition, SDKs, or pricing information for calls.
• The generation, detection, file handling, and payment requests that take place within a web page do not mean that external programs are allowed to be integrated.
• At present, there is no confirmed official open-source repository for the Miaoxie product, nor any product-level open-source license.
• The use of third-party front-end components or open technologies does not imply that the paper’s models, detection algorithms, and datasets are open source.

Frequently Asked Questions

How long can a paper be generated using Kaanran Miaoxie?

For current paper writing, three length options are available: 3000–5000 words, 5000–8000 words, and 8000–10000 words; the final actual word count still needs to be verified according to the requirements set by the university.

Is it possible to support undergraduate and master’s theses?

The page offers options for both undergraduate and master’s levels of education, which are used to adjust the way the content is presented; however, it is not guaranteed that the required standards for those respective degrees will be met automatically.

Can a literature review be generated?

The entry for an independent literature review is still marked as upcoming and should not be considered a feature that is already available.

Are the AIGC detection results consistent with those from CNKI and Wanfang?

It cannot be guaranteed; due to differences in models, thresholds, and versions across various systems, the same text may receive different scores.

How much quota is consumed by AI rewriting?

Both the rewriting itself and the subsequent verification checks are counted toward the character limit; it is necessary to check the estimated amount that will be deducted from the task before submitting it.

What files are supported by the format check?

Currently, only DOCX files up to 20MB in size are supported, and the output includes a corrected version as well as a version with annotations in Word format.

Can Kaoan Miaoxie be used for free?

The page indicates that a free trial of 2,000 characters is available; the specific functions offered, the eligibility requirements for new users, and the duration of the promotion are as indicated in the account details.

Does Ke’an Miao Write offer a public API or an open-source version?

Currently, there are no confirmed public developer APIs, official SDKs, open-source repositories, or product-level open-source licenses available.

How to add a FAQ Schema

This page uses level 3 headings to present the questions, with the corresponding answers provided right next to each paragraph; the website can retrieve these question-and-answer pairs on the server side to generate structured FAQ data.

Structured data should be consistent with the main content of the page; it should not include features such as hidden pass rates, detection guarantees, prices, ratings, links, or literature reviews for articles that have not yet been published.

Summary

Sholingo offers an excellent solution by combining the initial draft of a paper, Chinese-language AIGC tools, and Word formatting checks within the same web interface, making it useful for improving the structure, expression, and layout of texts.

The true quality of a paper still depends on genuine research, reliable sources, transparent disclosure, and author review; before purchasing a package, it is also necessary to check the pricing based on word count, the number of downloads allowed, as well as the policies regarding privacy and refunds.