- The webservice for the YODIE (previously LODIE) entity disambiguation system and associated datasets of annotated tweets (https://gate.ac.uk/applications/yodie.html) . See for detailed description and use the deliverables D2.2.1, D.2.2.2 and D2.3.2.
- A Twitter POS tagger developed by USFD is available here:
- The code associated with deliverable D3.1.1 Regression models of trends in streaming data is available at github: https://github.com/sinjax/trendminer. This includes also a preprocessing pipeline working as a standalone tool, Hadoop tool, or as a webservice.
- The code associated with deliverable D3.2.1 Clustering models for discovery of regional and demographic variation: https://github.com/danielpreotiuc/trendminer-clustering , now with pre-trained topic for 1 year of data with cluster coherence and labels provided by the partner SORA; released as a webservice.
- A data collection utilities for streaming collection, user-based collection, etc.: https://github.com/danielpreotiuc/twitter-collection-utils -
- Merging of (German) Polarity lexicons: https://github.com/guyemerson/SentiMerge. See also deliverable D2.3.2
The TrendMiner Ontologies, associated with the eliverables D2.2.2 (Multilingual, ontology-based IE from stream media -v2) and D2.1.2 (Knowledge & Provenance Modelling and Stream Reasoning - v2) : http://www.dfki.de/lt/onto/
- Demo of the Summarisation prototype for TrendMiner: http://demos.gate.ac.uk/trendminer/summarization/ Related to D.4.1.2 (: D4.1.2.pdf)
- Code for the BGGR model introduced in the deliverable D3.3.1. Please go to https://github.com/sinjax/trendminer-python/tree/master/bivariate
- Spatio-temporal clustering: Code to transform the data using RBF kernels for use with the DMR topic model from Mallet. Method is introduced in the deliverable D3.3.1. Please go to https://github.com/andreavarga/trendminer-sptempclustering
- Hungarian NLP pipeline for social media text analysis: https://github.com/mmihaltz/trendminer-hunlp
- Sentipejd – a sentiment analysis tool for Polish made available as a Web service in Multiservice framework for chaining NLP Web services
- Facebook data download, Java NooJ import/export format conversion: https://github.com/mmihaltz/trendminer-hutools