{
  "id": 76339,
  "title": "Sentiment Analysis For Stock Market",
  "url": "/competitions/human-protein-atlas-image-classification/discussion/76339",
  "author_name": "",
  "post_date": "2019-01-01T14:27:43.004477400Z",
  "votes": -19,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Hi \nDo you have an advise about model  to apply for  Sentiment Analysis for stock market.\nI tried to apply <a href=\"https://github.com/adeshpande3/LSTM-Sentiment-Analysis.git\">https://github.com/adeshpande3/LSTM-Sentiment-Analysis.git</a> that is based on movies review without much success</p>\n\n<p>Thanks\nSasha</p>",
  "messages": [
    {
      "id": "448571",
      "postDate": "01/01/2019 14:27:43",
      "content": "<p>Hi \nDo you have an advise about model  to apply for  Sentiment Analysis for stock market.\nI tried to apply <a href=\"https://github.com/adeshpande3/LSTM-Sentiment-Analysis.git\">https://github.com/adeshpande3/LSTM-Sentiment-Analysis.git</a> that is based on movies review without much success</p>\n\n<p>Thanks\nSasha</p>",
      "rawMarkdown": "Hi \nDo you have an advise about model  to apply for  Sentiment Analysis for stock market.\nI tried to apply https://github.com/adeshpande3/LSTM-Sentiment-Analysis.git that is based on movies review without much success\n\nThanks\nSasha",
      "votes": null
    },
    {
      "id": "448579",
      "postDate": "01/01/2019 15:08:29",
      "content": "<p>You'd be better off trying one of these: <a href=\"http://www.shamanscrystal.co.uk/category/5/crystal-balls/\">http://www.shamanscrystal.co.uk/category/5/crystal-balls/</a> ;)</p>",
      "rawMarkdown": "You'd be better off trying one of these: http://www.shamanscrystal.co.uk/category/5/crystal-balls/ ;)",
      "votes": null
    },
    {
      "id": "448659",
      "postDate": "01/01/2019 19:32:11",
      "content": "<p>Not sure this question is posted in the correct discussion board, but you <a href=\"https://github.com/maxbbraun/trump2cash\">might find this interesting</a>. </p>\n\n<p>To answer your question more fully, it might be helpful to know a bit more about what text sources you want to analyse and how you plan to use the output. As a simple example, you could scrape some business news web pages, tokenise the text and come up with a positive / negative score by comparing word use to the Bing Lexicon, and make buy / sell recommendations based on those scores. All that really involves is some simple data wrangling, some addition and calculating a net positive / negative score, no complicated models required.</p>\n\n<p>That might be a bit simplistic for your purposes, but, like I said, a bit more background and context might get some more useful answers than my basic suggestion!</p>",
      "rawMarkdown": "Not sure this question is posted in the correct discussion board, but you [might find this interesting][1]. \n\nTo answer your question more fully, it might be helpful to know a bit more about what text sources you want to analyse and how you plan to use the output. As a simple example, you could scrape some business news web pages, tokenise the text and come up with a positive / negative score by comparing word use to the Bing Lexicon, and make buy / sell recommendations based on those scores. All that really involves is some simple data wrangling, some addition and calculating a net positive / negative score, no complicated models required.\n\nThat might be a bit simplistic for your purposes, but, like I said, a bit more background and context might get some more useful answers than my basic suggestion!\n\n[1]: https://github.com/maxbbraun/trump2cash",
      "votes": null
    },
    {
      "id": "449075",
      "postDate": "01/02/2019 15:52:41",
      "content": "<p>Hi Chris\nI would  to classify yahoo finance news headlines for positive , negative and neutral without using Google's Cloud Natural Language API</p>\n\n<p>Thanks\nSasha</p>",
      "rawMarkdown": "Hi Chris\nI would  to classify yahoo finance news headlines for positive , negative and neutral without using Google's Cloud Natural Language API\n\nThanks\nSasha",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 448579,
      "author_name": "maw501",
      "author_url": "",
      "post_date": "01/01/2019 15:08:29",
      "content": "<p>You'd be better off trying one of these: <a href=\"http://www.shamanscrystal.co.uk/category/5/crystal-balls/\">http://www.shamanscrystal.co.uk/category/5/crystal-balls/</a> ;)</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 448659,
      "author_name": "chrisbow",
      "author_url": "",
      "post_date": "01/01/2019 19:32:11",
      "content": "<p>Not sure this question is posted in the correct discussion board, but you <a href=\"https://github.com/maxbbraun/trump2cash\">might find this interesting</a>. </p>\n\n<p>To answer your question more fully, it might be helpful to know a bit more about what text sources you want to analyse and how you plan to use the output. As a simple example, you could scrape some business news web pages, tokenise the text and come up with a positive / negative score by comparing word use to the Bing Lexicon, and make buy / sell recommendations based on those scores. All that really involves is some simple data wrangling, some addition and calculating a net positive / negative score, no complicated models required.</p>\n\n<p>That might be a bit simplistic for your purposes, but, like I said, a bit more background and context might get some more useful answers than my basic suggestion!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 449075,
      "author_name": "sashasofa",
      "author_url": "",
      "post_date": "01/02/2019 15:52:41",
      "content": "<p>Hi Chris\nI would  to classify yahoo finance news headlines for positive , negative and neutral without using Google's Cloud Natural Language API</p>\n\n<p>Thanks\nSasha</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "448571": "Hi \nDo you have an advise about model  to apply for  Sentiment Analysis for stock market.\nI tried to apply https://github.com/adeshpande3/LSTM-Sentiment-Analysis.git that is based on movies review without much success\n\nThanks\nSasha",
    "448579": "You'd be better off trying one of these: http://www.shamanscrystal.co.uk/category/5/crystal-balls/ ;)",
    "448659": "Not sure this question is posted in the correct discussion board, but you [might find this interesting][1]. \n\nTo answer your question more fully, it might be helpful to know a bit more about what text sources you want to analyse and how you plan to use the output. As a simple example, you could scrape some business news web pages, tokenise the text and come up with a positive / negative score by comparing word use to the Bing Lexicon, and make buy / sell recommendations based on those scores. All that really involves is some simple data wrangling, some addition and calculating a net positive / negative score, no complicated models required.\n\nThat might be a bit simplistic for your purposes, but, like I said, a bit more background and context might get some more useful answers than my basic suggestion!\n\n[1]: https://github.com/maxbbraun/trump2cash",
    "449075": "Hi Chris\nI would  to classify yahoo finance news headlines for positive , negative and neutral without using Google's Cloud Natural Language API\n\nThanks\nSasha"
  },
  "source": "meta"
}