0 ×

24_​Analyzing_​Twitter_​Posts_​with_​Custom_​Tagging

Workflow

KNIME Tweet Analysis
text processingNLPNatural Language Processing
Textprocessing on KNIME tweetsThis workflow applies textprocessing on KNIME tweets (i.e. tweets containing the string KNIME). To save you the hassle of requesting Twitter API credentials, we stored some KNIMEtweets for you in a H2 Database within this workflow. The intention of this workflow is to showcase the Dictionary Tagger (Multi Column) and Term Neighborhood Extractor nodes,which come as new features with the release of KNIME Analytics Platform 3.6. The tweets are tagged using two overlapping dictionaries, one containing data science terms and the other containing terms related to AI. The tagged terms are displayed in a tag cloudwhere the number of users who used that terms corresponds to the size of the term. To investigate if the displayed tweets have a positive or negative connotation, we did a sentimentanalysis on the term neighborhood and displayed that in an interactive table. Read tweets from DB Read dictionaries Tag and count all occurences from the data science/ML dictionaries in KNIME tweets Extract neighborhoods of data science/AI terms and do sentiment analysis on them Create interactive view Additional informationTwitter sentiment dictionary: M. Ghiassi, J. Skinner, D. Zimbra (2013): Twitter brand sentiment analysis: A hybrid system using n-gram analysis and dynamic artificial neural network, in Expert Systems with Applications, Volume 40, Issue 16, Pages 6266-6282,available at http://kt.ijs.si/markodebeljak/Lectures/Seminar_MPS/2012_on/Seminars_2015_16/Simon%20Brmez/Bibliography/%5B20%5D%20Twitter%20brand%20sentiment%20analysis%20A%20hybrid%20system%20using%20n-gram%20analysis%20and%20dynamic%20artificial%20neural%20network.pdf (downloaded on June 20th, 2018)Data Science dictionary: B. DuCharme (2018): data science glossary, available at http://www.datascienceglossary.org/ (requestedon June 20th, 2018)Machine Learning dictionary: Google (2018): Machine Learning Glossary, available at https://developers.google.com/machine-learning/glossary/ (requested on June 20th, 2018) Create documents from TweetsTag from both dictionariesBag em up!Filter Group termson unique countsTo upper caseString termsTag from both dictionariesJoin combined data science and ML dictionaries withpositiv/negative dictionariesNeighborhoodof oneTo upper caseFilterselect relevant tableread KNIMErelated TweetsStrings To Document Dictionary Tagger(Multi Column) Bag Of WordsCreator Tag Filter Tags To String GroupBy Case Converter GroupBy Term To String Column Combiner String Manipulation Dictionary Tagger(Multi Column) Joiner Term NeighborhoodExtractor Case Converter Tag Filter Joiner Tag Cloud &Table View Post processingfor view Read ML and DataScience dictionaries Read Twitter Sentimentdictionaries Create sentimentscore DB Table Selector DB Reader H2 Connector Textprocessing on KNIME tweetsThis workflow applies textprocessing on KNIME tweets (i.e. tweets containing the string KNIME). To save you the hassle of requesting Twitter API credentials, we stored some KNIMEtweets for you in a H2 Database within this workflow. The intention of this workflow is to showcase the Dictionary Tagger (Multi Column) and Term Neighborhood Extractor nodes,which come as new features with the release of KNIME Analytics Platform 3.6. The tweets are tagged using two overlapping dictionaries, one containing data science terms and the other containing terms related to AI. The tagged terms are displayed in a tag cloudwhere the number of users who used that terms corresponds to the size of the term. To investigate if the displayed tweets have a positive or negative connotation, we did a sentimentanalysis on the term neighborhood and displayed that in an interactive table. Read tweets from DB Read dictionaries Tag and count all occurences from the data science/ML dictionaries in KNIME tweets Extract neighborhoods of data science/AI terms and do sentiment analysis on them Create interactive view Additional informationTwitter sentiment dictionary: M. Ghiassi, J. Skinner, D. Zimbra (2013): Twitter brand sentiment analysis: A hybrid system using n-gram analysis and dynamic artificial neural network, in Expert Systems with Applications, Volume 40, Issue 16, Pages 6266-6282,available at http://kt.ijs.si/markodebeljak/Lectures/Seminar_MPS/2012_on/Seminars_2015_16/Simon%20Brmez/Bibliography/%5B20%5D%20Twitter%20brand%20sentiment%20analysis%20A%20hybrid%20system%20using%20n-gram%20analysis%20and%20dynamic%20artificial%20neural%20network.pdf (downloaded on June 20th, 2018)Data Science dictionary: B. DuCharme (2018): data science glossary, available at http://www.datascienceglossary.org/ (requestedon June 20th, 2018)Machine Learning dictionary: Google (2018): Machine Learning Glossary, available at https://developers.google.com/machine-learning/glossary/ (requested on June 20th, 2018) Create documents from TweetsTag from both dictionariesBag em up!Filter Group termson unique countsTo upper caseString termsTag from both dictionariesJoin combined data science and ML dictionaries withpositiv/negative dictionariesNeighborhoodof oneTo upper caseFilterselect relevant tableread KNIMErelated TweetsStrings To Document Dictionary Tagger(Multi Column) Bag Of WordsCreator Tag Filter Tags To String GroupBy Case Converter GroupBy Term To String Column Combiner String Manipulation Dictionary Tagger(Multi Column) Joiner Term NeighborhoodExtractor Case Converter Tag Filter Joiner Tag Cloud &Table View Post processingfor view Read ML and DataScience dictionaries Read Twitter Sentimentdictionaries Create sentimentscore DB Table Selector DB Reader H2 Connector

Download

Get this workflow from the following link: Download

Nodes

24_​Analyzing_​Twitter_​Posts_​with_​Custom_​Tagging consists of the following 51 nodes(s):

Plugins

24_​Analyzing_​Twitter_​Posts_​with_​Custom_​Tagging contains nodes provided by the following 6 plugin(s):