english  login  

appstore
  • [Automatic Plug-in] Topic Modeling

  • Author : admin Version : 1.0 Date : 2018-06-01 17:09:02
  • link copyURL 복사
  • Author : CYRAM
    Date : 2018-03-16
    Version : 1.4 (2018-06-01. It enables you to run 'Words in the original text' on Topic-Word network map)
    Description : This plug-in is useful to find the hidden topics and visualize the result of topic modeling(LDA)
    Some newest update features of NetMiner are included in these plug-ins, please install the latest version of NetMiner(4.4.0.b).

    ----- How to Install ------

    1) Download an attachment.
    2) Register the NMX file in the plug-in manager.

    ----- How to Use ------

    * This plug-in requires 'Semantic Network Edition' or higher license.

    ① Import unstructured text data before running this plug-in(Menu File >> Import unstructured text)
    ② Go to Tools > Plug-ins > '[Automatic Plug-in] Topic Modeling'
    ③ Enter the threshold values of TF-IDF and word length to delete useless words.
    ④ Enter values for topic modeling(LDA) and visualization between topic and main words.
    ⑤ Press 'OK' button.
    ⑥ The process will be started. When the process is done, the following results will be generated.
    - Filtered Words : Words excluding the common and short words
    - Result[Documents...] : Word networks for each topic(You can find a network map for each topic by double-clicking 'Spring 2D' in sessions)
    - Topic proportion : Pie chart of 'Frequency' in sessions shows the number of documents for each topic.
    - Topic map : Visualization for word - Topic network

    For more description about this plug-in, please visit following URL.
    (Eng) http://www.netminer.com/community/forum/forum-view.do?wr_id=232
    (Kor) http://cyram.tistory.com/249?category=491192

  • list