Your data collection violates copyright laws. Downloading articles from SciHub is illegal. Typically, no publisher will come after you for downloading a few articles, but you are proposing to download millions of articles, and then to advertise to the world what you did by publishing an article about it.

Do not do this.

First, most publishers will not accept a paper based on illegal data collection. Commercial publishers won't like it because you are... demonstrating your contempt for their business model. (I toyed with other metaphors.) And even if "nonstandard" publishers might like it, e.g., journals run entirely by open science aficionado academics whose philosophy and ethics this very much resonates with, if they have any sense at all, they will not publish something like your paper, because it opens them up to serious litigation. And they know full well that the commercial publishers will absolutely love having an opening like this to put them out of business, and at the same time send a clear message to all the other academics that might consider intruding on their turf.

Because, second, make no mistake: if you plan on doing something this blatantly illegal, and publishing a paper about it, then litigation may well be coming your way. There have been high profile cases about people breaking copyright in similar ways, and publishers have enough legal experts to drag you through courts for years. (Edit - Franck Dernoncourt asks the obvious follow-up question here: Has any researcher been sued for using Sci-Hub data in their research?) And at the same time, your internet service provider will get wind of this and start litigating as well, because you very probably broke your service agreement in multiple ways. (Did you read all the fine print you accepted?) If you are doing this from a university account, the least that will happen is that your IT department will close your account, then you will be expelled if you are a student, and terminated if you are an employee. Kiss your hopes for a degree goodbye.

If you write a paper that actually gets published and nobody reads, so the legal s***shorm does not arrive immediately, you are now living with a huge problem on your hands that could explode into your face at absolutely any time in the rest of your life. You do not know whether someone will dig this up in ten years and make your life miserable in the ways described above, even if the legal aspects are under a statute of limitations. You don't know how copyright laws and court decisions will develop in the coming decades, but doing something like this now is not a good starting point for the rest of your life.

Do not do this.

The data collection you describe probably violates copyright laws. The specific legality of downloading articles from SciHub depends on the jurisdiction and on other circumstances; e.g., personal use is legal, but scientific research is not in Switzerland, and your proposed scraping of millions of articles will probably rather fall in the latter category. Typically, no publisher will come after you for downloading a few articles, but you are proposing to download millions of articles, and then to advertise to the world what you did by publishing an article about it.

First, most publishers will not accept a paper based on questionable-to-illegal data collection. (You might be able to argue that your particular approach was indeed legal in your jurisdiction, but then things become hairy if the jurisdiction of the publisher is also relevant, and nobody really wants to get their legal team involved in the publishing decision.) Commercial publishers won't like it because you are... demonstrating your contempt for their business model. (I toyed with other metaphors.) And even if "nonstandard" publishers might like it, e.g., journals run entirely by open science aficionado academics whose philosophy and ethics this very much resonates with, if they have any sense at all, they will not publish something like your paper, because it opens them up to serious litigation. And they know full well that the commercial publishers will absolutely love having an opening like this to put them out of business, and at the same time send a clear message to all the other academics that might consider intruding on their turf.

Because, second, make no mistake: if you plan on doing something this (probably) illegal, and publishing a paper about it, then litigation may well be coming your way. There have been high profile cases about people breaking copyright in similar ways, and publishers have enough legal experts to drag you through courts for years. (Edit - Franck Dernoncourt asks the obvious follow-up question here: Has any researcher been sued for using Sci-Hub data in their research?) And at the same time, your internet service provider will get wind of this and may start litigating as well, because you very probably broke your service agreement in multiple ways. (Did you read all the fine print you accepted?) If you are doing this from a university account, the least that can happen is that your IT department could close your account, then you may be expelled if you are a student, and terminated if you are an employee.

If you write a paper that actually gets published and nobody reads, so the legal s***shorm does not arrive immediately, you are now living with a huge problem on your hands that could explode into your face at absolutely any time in the rest of your life. You do not know whether someone will dig this up in ten years and make your life miserable in the ways described above, even if the legal aspects are under a statute of limitations. You don't know how copyright laws and court decisions will develop in the coming decades, but doing something like this now is not a good starting point for the rest of your life.

And all of these may well happen even if you have a case that your method was legal in the first place. If your school expels you over this and you win a case against it in court in five years, then yes, you did win, but was the legal hassle worth it?

Your data collection violates copyright laws. Downloading articles from SciHub is illegal. Typically, no publisher will come after you for downloading a few articles, but you are proposing to download millions of articles, and then to advertise to the world what you did by publishing an article about it.

Do not do this.

First, most publishers will not accept a paper based on illegal data collection. Commercial publishers won't like it because you are... demonstrating your contempt for their business model. (I toyed with other metaphors.) And even if "nonstandard" publishers might like it, e.g., journals run entirely by open science aficionado academics whose philosophy and ethics this very much resonates with, if they have any sense at all, they will not publish something like your paper, because it opens them up to serious litigation. And they know full well that the commercial publishers will absolutely love having an opening like this to put them out of business, and at the same time send a clear message to all the other academics that might consider intruding on their turf.

Because, second, make no mistake: if you plan on doing something this blatantly illegal, and publishing a paper about it, then litigation is coming your way. There have been high profile cases about people breaking copyright in similar ways, and publishers have enough legal experts to drag you through courts for years. And at the same time, your internet service provider will get wind of this and start litigating as well, because you very probably broke your service agreement in multiple ways. (Did you read all the fine print you accepted?) If you are doing this from a university account, the least that will happen is that your IT department will close your account, then you will be expelled if you are a student, and terminated if you are an employee. Kiss your hopes for a degree goodbye.

If you write a paper that actually gets published and nobody reads, so the legal s***shorm does not arrive immediately, you are now living with a huge problem on your hands that could explode into your face at absolutely any time in the rest of your life. You do not know whether someone will dig this up in ten years and make your life miserable in the ways described above, even if the legal aspects are under a statute of limitations. You don't know how copyright laws and court decisions will develop in the coming decades, but doing something like this now is not a good starting point for the rest of your life.

Do not do this.

Your data collection violates copyright laws. Downloading articles from SciHub is illegal. Typically, no publisher will come after you for downloading a few articles, but you are proposing to download millions of articles, and then to advertise to the world what you did by publishing an article about it.

Do not do this.

First, most publishers will not accept a paper based on illegal data collection. Commercial publishers won't like it because you are... demonstrating your contempt for their business model. (I toyed with other metaphors.) And even if "nonstandard" publishers might like it, e.g., journals run entirely by open science aficionado academics whose philosophy and ethics this very much resonates with, if they have any sense at all, they will not publish something like your paper, because it opens them up to serious litigation. And they know full well that the commercial publishers will absolutely love having an opening like this to put them out of business, and at the same time send a clear message to all the other academics that might consider intruding on their turf.

Because, second, make no mistake: if you plan on doing something this blatantly illegal, and publishing a paper about it, then litigation may well be coming your way. There have been high profile cases about people breaking copyright in similar ways, and publishers have enough legal experts to drag you through courts for years. (Edit - Franck Dernoncourt asks the obvious follow-up question here: Has any researcher been sued for using Sci-Hub data in their research?) And at the same time, your internet service provider will get wind of this and start litigating as well, because you very probably broke your service agreement in multiple ways. (Did you read all the fine print you accepted?) If you are doing this from a university account, the least that will happen is that your IT department will close your account, then you will be expelled if you are a student, and terminated if you are an employee. Kiss your hopes for a degree goodbye.

If you write a paper that actually gets published and nobody reads, so the legal s***shorm does not arrive immediately, you are now living with a huge problem on your hands that could explode into your face at absolutely any time in the rest of your life. You do not know whether someone will dig this up in ten years and make your life miserable in the ways described above, even if the legal aspects are under a statute of limitations. You don't know how copyright laws and court decisions will develop in the coming decades, but doing something like this now is not a good starting point for the rest of your life.

Do not do this.