About Sentiment Analysis and the analysis method
Sentiment Analysis belongs to the research area of Text Mining. Sentiment Analysis explores the polarity of a text, sentence or token, differentiating between positive, neutral and negative polarity as well as their polarity strength. Words that have a sentiment are known as "sentiment bearing words". To identify these sentiment bearing words, we apply a sentiment dictionary - this is referred to as lexicon-based sentiment analysis. We check whether a certain token is in this lexicon and if yes, we use the polarity and polarity strength this token has for further calculation. To calculate the sentiment score of a specific text unit (e.g. the entire text, sentence), these polarity strengths of the corresponding level are combined. For more information about the lexicons we use per default and further research, please refer to the Documentation.
SentText - A Tool for Lexicon-Based Sentiment Analysis
SentText is a web application for sentiment analysis with a focus on digital humanities. The application offers numerous functions that support you in the analysis of your texts via lexicon based sentiment analysis. Our functions and lexicons currently focus on German sentiment lexicons and functionalites (because German texts are the texts we are primarly interested), but we plan to extend the functions to other languages as we continue the development. Please note that this application is still a prototype undergoing developement.
- Detailed presentation of the analysis
- Manual post-processing of the sentiment analysis
- Group files into folders
- Folders and file comparison
- Import of files and export of results
Demo with German lexicon (SentiWS)
The above paper was presented at the 16th International Symposium of Information Science (ISI 2021). Find more information about the conference here. You can access the entire proceedings here.
The presentation that was given about SentText at ISI 2021 can be found on Youtube.
- Mail: firstname.lastname@example.org