Rexha Andi, Dragoni Mauro , Federici Marco
An Unsupervised Aspect Extraction Strategy For Monitoring Real-Time Reviews Stream
One of the most important opinion mining research directions falls in the extraction ofpolarities referring to specific entities (aspects) contained in the analyzed texts. Thedetection of such aspects may be very critical especially when documents come fromunknown domains. Indeed, while in some contexts it is possible to train domainspecificmodels for improving the effectiveness of aspects extraction algorithms, inothers the most suitable solution is to apply unsupervised techniques by making suchalgorithms domain-independent and more efficient in a real-time environment. Moreover,an emerging need is to exploit the results of aspect-based analysis for triggeringactions based on these data. This led to the necessity of providing solutions supportingboth an effective analysis of user-generated content and an efficient and intuitive wayof visualizing collected data. In this work, we implemented an opinion monitoringservice implementing (i) a set of unsupervised strategies for aspect-based opinion miningtogether with (ii) a monitoring tool supporting users in visualizing analyzed data.The aspect extraction strategies are based on the use of an open information extractionstrategy. The effectiveness of the platform has been tested on benchmarks provided by the SemEval campaign and have been compared with the results obtained by domainad aptedtechniques.