- Bug fixed so Sentiment Analysis can be used also in Catalan, Portuguese and Italian with user defined models.
- The add-in now uses the Sentiment Analysis 2.1 and allows to use user sentiment models in your analysis.
- The MeaningCloud environment used can be configured so it can be used against on-premise solutions.
- A bug associated to the selection of combined cells has been fixed.
- Bug fixed in updates check with proxy connection.
- The add-in now uses the Topics Extraction 2.0.
- Text Clustering has beean added.
- The latest classification models published (Social Media and IAB) have been added to the Text Classification functionality.
- General usability has been improved by:
- An additional configurability option has been added to avoid combining cells in the output and provide flexibility in analyses over the results.
- Empty cells are now ignored instead of printing an error.
- Slightly modifying data input for IDs.
- The add-in remembers from on run to the next the last analysis settings used.
- We've added the sense ID of the entities and concepts detected in the Sentiment Analysis output.
- The relative relevance has been added as an output field in Text Classification.
- The Sense ID of entities and concepts has been added to the output of Topics Extraction.
- A bug in the feature-level analysis has been fixed (incorrect offset for texts without concepts or entities).
- The add-in now uses the Sentiment Analysis 2.0.
- Support for user dictionaries in Topics Extraction has been added.
- Issues with rate limit have been fixed.
- A new button to check if the add-in is up to date has been added.
Click on the version number to see the change log.