There have been several algorithms written which help identify authors based on their writing, researchers can determine a sex, culture, or psychological displacement by words in their works. Researchers can tell if the author or writer of a particular article or piece is angry, psychologically disturbed, or even depressed. These are obviously great tools for the Internet, especially when government agencies, law enforcement, or intelligence services are going through the hordes of data and information being put online each and every day.
What may appear to be a joke could be red flagged by a computer algorithm as a potential terrorist threat, even though it was just a teenager playing a trick or texting his buddies. Yes, this has happened before, or a couple of students were coming to the United States and they made some comments on twitter and Facebook about what they were going to do in the United States once they got here. They were no more terrorists than you are I, but they were stopped at LAX and refused admission into the country, denied entry.
Okay so, back to what I’m here today to talk about. Since there are several algorithms out there to identify authors, it appears to me that we can also design an algorithm to tell us when something was written based on various clues, uses of the language, as to when a specific research paper may have been written. One problem when searching in Google Scholar is that so many research papers do not have dates on them. Interestingly enough, all the older research paper do but many of the new scientific papers do not. Why is this you ask?
Well, sometimes it is because the paper was written for a symposium or a journal and bundled together with many other papers, but then unbundled for the search engine usage. The date no longer remains with that paper. Still, one way to find out when the paper was written is to go to the reference section and look at the dates of all the papers that it had referenced, realizing that it cannot reference a date to a paper which hasn’t been written yet, so the paper you are viewing has to be written after each and every one of the references, not before.
It would seem to me that Google Scholar needs an algorithm to determine the date of a research paper within a close proximity. I believe this can be done merely by the information in the paper. If it references information which has not been created yet, then that particular research paper is of more value, but it also can’t reference information that has not been written yet, or is not known yet by the scientific community.
Once these artificial intelligence algorithms can understand what is written and when the various concepts, theories, and concepts had come into fruition, it can more easily figure out the date of the paper. A human researcher can do this also, just by being well-versed in the subject matter, as a matter fact if I am studying something I’m pretty familiar with, I could pretty much pinpoint when a paper was written based on who wrote it, and what level of research we are at by looking at the various sources and data presented within six months to one year.
Sometimes you can determine with 90% probability within three months of when it was actually written. Of course, that takes some research in and of itself, and that’s why it would be nice if Google Scholar did indeed have such an algorithm. If you’d like to discuss this at much higher level, you may shoot me an e-mail. Until then I hope you will please consider all this and think on i