math_of_money
Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Application of text analytics
08-08-2018, 08:04 AM. (This post was last modified: 08-08-2018, 08:06 AM by gzbkel.)
Post: #1
Application of text analytics
Interesting article on application of text analytics on employee monitoring and financial analysis.

...Text analytics has become especially popular in finance. Investment banks and hedge funds scour public filings, corporate press releases, and statements by executives to find slight changes in language that might indicate whether a company’s stock price is likely to go up or down; Goldman Sachs calls this kind of natural-language processing “a critical tool for tomorrow’s investors.” Specialty-research firms use artificial-intelligence algorithms to derive insights from earnings-call transcripts, broker research, and news stories...


https://www.theatlantic.com/magazine/arc...ource=feed

Find Reply
08-08-2018, 09:14 AM. (This post was last modified: 08-08-2018, 09:15 AM by opmi.)
Post: #2
RE: Application of text analytics
I guess it is easier for SEC US and HKEX Announcements. More text based.

SGX annc website messy and multi-layers. Not web scraping friendly. Parsing PDFs can be a pain in the ass. Cleaning up data is painful too.

Hope SGX change to more HK type.

I Guess the secret sauce in financial NLP is in developing the corpus.
Like some folks train European languages translations using EU govt docs.
"... but quitting while you're ahead is not the same as quitting." - Quote from the movie American Gangster

Find Reply
08-08-2018, 10:00 AM.
Post: #3
RE: Application of text analytics
(08-08-2018, 09:14 AM)opmi Wrote: I guess it is easier for SEC US and HKEX Announcements. More text based.

SGX annc website messy and multi-layers. Not web scraping friendly. Parsing PDFs can be a pain in the ass.  Cleaning up data is painful too.

Hope SGX change to more HK type.

I Guess the secret sauce in financial NLP is in developing the corpus.
Like some folks train European languages translations using EU govt docs.

Yes, I agree it is very messy.
Maybe they design the SGX website to deter scraping. But it should be possible to scrape the links to all financial reports PDF files (For example, like Disclosure of Interest in the VB site).
PDF files can be read using free libraries, but it will be one big chunk of unstructured text. Need to search for keywords like "Chairman's message" etc, but I think still can be done.

It's the NLP part which is hard and I have no knowledge about.

Find Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)
Valuebuddies.com | Return to Top | Lite (Archive) Mode | RSS Syndication | CONTACT US: nas......@valuebuddies.com | | Share Buy-Back | Disclosure of Interest