TourPedia is a Web application which shows users’ sentiments about touristic locations of some of the most important cities and regions in Europe. More specifically, TourPedia, exploits the OpeNER pipeline to analyse users’ reviews on places. All reviews are extracted from social media. Once analysed, each review is associated to a rate, which ranges from 0 to 100. The sentiment of each place is calculated as a function of all the sentiments of reviews on that place. As a result, TourPedia shows all the places and their users’ sentiments on a map.
EntitUp is a Social and Media Analysis and Monitoring application. It provides Media Monitoring solutions on User-Generated Contents (UCGs). The user can choose to analyze both online Travel Books and online Cities Documentaries. The statistics focus on Named Entity Recognition and Sentiment Analysis and provide the user with quantitative and qualitative information.
KAF Browser is a demo that shows some of the OpeNER tools analysis results. The demo shows many views of the information contained in a KAF documents, like the named entities detected (people, locations, organizations, etc.), their links to DBpedia if any, the polarity of the words, etc. The demo does not cover all the OpeNER analysis modules but it is a quick way to see what OpeNER project is about.
Moodmap is a resulting application of the second Hackathon in Amsterdam. The application analysed all the tweets of LREC (taking in account the #LREC2014 hashtag, and poster and oral sessions number specified in the tweet) and analysed and show alive the results twits and the mood of them. This demo had a huge interest in LREC2014 and a lot of participants tweet and came to our OpeNER booth to see the output analysed.
The Review Browser is a demo application that demonstrates what someone can potentially do by using the OpeNER tool chain in the Hotel domain. Reviews are becoming more and more important in the hotel industry, since someone can draw useful conclusions about several aspects of the hotel from the customer perspective. Apart from the ratings of a review, there is much more information hidden in the context. That is where OpeNER gets involved. Given that each hotel review is processed by the whole OpeNER chain, (Language identifier, tokenizer, POS tagger, polarity tagger, property tagger, constituent parser, ner, co-reference, ned, opinion detector with hotel domain models) they are then stored in a database, in a way where all the needed information is stored as well. The most important information is the property tags and the opinion(s) of each review. Each review is analyzed and it is known what the opinion of each customer was for each domain that he was referring in his review. That way, not only important information is extracted from a review, which would otherwise be ignored or would require manual work to be received, someone can also view related reviews that refer to the same topic, assign a task to a specific person, regarding the review or write a note about it.
Link coming soon…