Mobile review dataset for aspect level sentiment analysis

It is a dataset containing sentence segments from cutomer reviews about mobile phone from different sources like Amazon, Flipkart, Tweeter and some existing datasets. It contains more than 1000 records tagged with one of the five aspect categories battery, camera, display, price and processor. Whether a sentence segment has sentiment expression (subjective/ objective) is also tagged and the sentiment orientation (positive/ negative/ neutral) of each sentence segment is assigned. Explicit or implicit presence of aspect is also maintained. It is a balanced dataset containing at least 200 records for each aspect category. Balance between explicit and implicit presence of aspect is also maintained by having at least 80 records of each type for every aspect category.