Wikipedia:Reference desk/Archives/Computing/2017 June 13

Computing desk
< June 12 << May | June | Jul >> June 14 >
Welcome to the Wikipedia Computing Reference Desk Archives
The page you are currently viewing is a transcluded archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.


June 13

edit

Internet Security and SSN

edit

I have read on websites that advise against inserting the social security number too early on in the job application. Usually, a legitimate employer should request the SSN later in the application process, like after offering the job to the candidate. Similarly, paper applications (from Panera Bread) and electronic kiosks (from Target) may ask for SSN. Why do employers ask for SSN so soon in the first place, right on the job application? For security purposes, can one make up a SSN, and if the SSN is later required for a background check, then the real SSN is offered? Or is it best to stop applying to those corporations altogether? SSN identity theft seems to be the highest kind of theft, as it is attached to bank accounts, credit scores, and medical history. Why can't all legitimate employers by law ask for SSN after the hiring? Just how common is SSN theft? 50.4.236.254 (talk) 11:14, 13 June 2017 (UTC)[reply]

It appears that unless you live in New York, Connecticut or Massachusetts you don't have a choice if you want a good shot at getting the job. These two links are quite informative: 12. Thanks Jenova20 (email) 11:36, 13 June 2017 (UTC)[reply]
One reason to ask for it early is to discourage illegal immigrants from applying, as they won't have one. Of course, they can make one up, but must know it will be checked at some time, making it pointless to even apply. And yes, there are other ways potential employers can verify citizenship, but they cost time and money. If illegal immigrants never apply for jobs in the first place, then they don't waste either. StuRat (talk) 15:00, 14 June 2017 (UTC)[reply]

Is there a tool that can 'Twitterize' text to be fit 140 characters?

edit

Is there a tool that can take a piece of text and intelligently shorten it such that it fits Twitter's 140 character restriction? I'm aware of URL shorteners, and I am using one, but I'm looking for something more, something that can parse text and remove unnecessary words and replace words with common abbreviations. A Quest For Knowledge (talk) 11:51, 13 June 2017 (UTC)[reply]

Yes, it's called the human brain. In all seriousness, unless there are very strict guidelines as to how the original text is composed, or what subject the text is about, it would be extremely difficult to get a computer to summarize text in a cogent way. Now, if you just want something that can find and replace certain words and phrases, something like this might be what you're looking for. I've never used it before, so I can't give an opinion on it's usefulness or ease of use, though. ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it. 13:08, 13 June 2017 (UTC)[reply]
Sure, nobody said it was easy. But the world is full of people who like to do difficult things with computers, and sometimes let other people freely use the things they make. See here [1] for examples of how the routines I linked below are used productively all over reddit on a daily basis. Note: of course it's not as good as a human, and sometimes it fails miserably. But often this can take in general text and summarize it in a useful manner. SemanticMantis (talk) 17:32, 13 June 2017 (UTC)[reply]
Could you could use the text as an image instead? People are always doing that when reposting screenshots and walls of text from elsewhere. Thanks Jenova20 (email) 14:40, 13 June 2017 (UTC)[reply]
Sounds like you want something similar to a Automatic summarization tool or a text simplification tool. However, usually they're used for summarizing documents or paragraphs into a handful of sentences. I'm afraid I'm not aware of one that works well on individual sentences. ApLundell (talk) 16:12, 13 June 2017 (UTC)[reply]
However, You may be interested in Twitterative and Tweet Compressor, and this interesting tool.
ApLundell (talk) 16:18, 13 June 2017 (UTC)[reply]
This page [2] summarizes text using language processing algorithms. SemanticMantis (talk) 17:32, 13 June 2017 (UTC)[reply]
You could auto-disemvowel the text (for which apps exist) and then restore the few vowels necessary to remove ambiguity. Before disemvowelling's use for internet forum moderation, similar abjad modifications of written English were used (manually) as one form of business shorthand which, obviating the need to learn new symbols, was quicker to learn though not ultimately as fast as symbolic shorthands. (One used to see ads in London Underground carriages offering secretarial courses – "F u cn rd ths u cn bcm a scrtry & gt a gd jb.") {The poster formerly known as 87.81.230.195} 2.217.208.38 (talk) 12:26, 14 June 2017 (UTC)[reply]
Thanks, there seems to be an active area of development for these types of tools. I've tried a number of them, they all seem to do things differently, and there is no one stand out that I tried. But thanks, I'll just use my human brain for now. This does seem like a good candidate for an open source code project. A Quest For Knowledge (talk) 15:04, 14 June 2017 (UTC)[reply]

How do I convert this into JavaScript?

edit

This is apgsearch and I am trying to understand the code, but I need it in JavaScript.

# *************************************
# * Ash Pattern Generator (apgsearch) *
# *************************************
# * Version: v1.1  (beta release)     *
# *************************************
...large program redacted....

# apg_verify()

32ieww (talk) 22:56, 13 June 2017 (UTC)[reply]

I have redacted the program, as it really is too big for here, and is probably not a compatible copyright license. The program should be available here: http://catagolue.appspot.com/binaries/apgsearch-2015-05-25.py . Do a google search for "python to javascript conversion" and see if there is anything that you like there. It will not be super straight forward as there are dependencies with libraries that you need to substitute with some other code as well. Also the data structure will ahve to be converted too somehow. Examples include Pijs[3] and transcrypt[4] or skulpt, brython, pyjaco and pythonScript[5]. Another alternative is to grab the C++ version and convert that. Graeme Bartlett (talk) 00:06, 14 June 2017 (UTC)[reply]