Facebook Reaction Prediction for Sinhala Text - FYP Project Final Presentation

The Facebook network allows its users to record their reactions to text via a typology of emotions. This network, taken at scale, is, therefore, a prime data set of annotated sentiment data. In this research, the relationship between Facebook posts and the corresponding reactions is examined. The project can be separated into four steps. In the first step, millions of such reactions, derived from a decade worth of Facebook post data centred around a Sri Lankan context are used to model an eye of the beholder approach to sentiment detection for online Sinhala textual content. Three different sentiment analysis models are built, taking into account a limited subset of reactions, all reactions, and another that derives a positive/negative star rating value. The analysis reveals that binary classification of reactions, for Sinhala content, is significantly more accurate than the other approaches. In the second step, we test state-of-the-art Sinhala sentiment analysis models against the Facebook data set for the purpose of establishing benchmarks and with the goal of identifying the best model for Sinhala sentiment analysis. The 3 layer Bidirectional LSTM model achieves an F1 score of 84.58% for Sinhala sentiment analysis, surpassing the current state-of-the-art model; Capsule B, which only manages to get an F1 score of 82.04%. In the process of mathematically modelling natural languages, developing language embeddings is a vital step. In the third step, we explore the effectivity of several one-tiered and two-tiered embedding architectures in representing Sinhala text in the sentiment analysis domain. With our findings, the two-tiered embedding architecture has been proven to perform better than one-tier word embeddings, by achieving a maximum F1 score of 88.04% in contrast to the 83.76% achieved by word embedding models. Furthermore, embeddings in the hyperbolic space are also developed compared with Euclidean embeddings in terms of performance. The fourth and final step is to develop a novel model that can be used in the sentiment analysis arena. With the conclusions of the previous steps, we have selected to develop a model that consists of the Capsule B model and the Bi-LSTM model.