The inherent nature of social media content poses serious challenges to practical applications of sentiment analysis. We present VADER, a simple rule-based model for general sentiment analysis, and compare its effectiveness to eleven typical state-of-practice benchmarks including LIWC, ANEW, the General Inquirer, SentiWordNet, and machine learning oriented techniques relying on Naive Bayes, Maximum Entropy, and Support Vector Machine (SVM) algorithms. Using a combination of qualitative and quantitative methods, we first construct and empirically validate a gold-standard list of lexical features (along with their associated sentiment intensity measures) which are specifically attuned to sentiment in microblog-like contexts. We then combine these lexical features with consideration for five general rules that embody grammatical and syntactical conventions for expressing and emphasizing sentiment intensity. Interestingly, using our parsimonious rule-based model to assess the sentiment of tweets, we find that VADER outperforms individual human raters (F1 Classification Accuracy = 0.96 and 0.84, respectively), and generalizes more favorably across contexts than any of our benchmarks.
Pouyan, Maziyar Baran (University of Texas at Dallas) | Yousefi, Rasoul (University of Texas at Dallas) | Ostadabbas, Sarah (University of Texas at Dallas) | Nourani, Mehrdad (University of Texas at Dallas)
Pattern classification algorithms have been applied in data mining and signal processing to extract the knowledge from data in a wide range of applications. The Fuzzy inference systems have successfully been used to extract rules in rule-based applications. In this paper, a novel hybrid methodology using: (i) fuzzy logic (in form of if-then rules) and (ii) a bio-inspired optimization technique (firefly algorithm) is proposed to improve performance and accuracy of classification task. Experiments are done using nine standard data sets in UCI machine learning repository. The results show that overall the accuracy and performance of our classification are better or very competitive compared to others reported in literature.
Widely adopted for more than 20 years in industrial fields, business rules offer the opportunity to non-IT users to define decision-making policies in a simple and intuitive way. When used conjointly with probabilistic graphical models (PGM) their expressiveness increase by introducing the notion of probabilistic production rules (PPR). In this paper we will present a new model for PPR and suggest a way to handle the combinatorial explosion due to the number of parents of aggregators in PGM such as Bayesian Networks and Probabilistic Relational Models in an industrial context where marginals should be computed rapidly.
Adverse Drug Reactions (ADRs) represent troublesome and potentially fatal side effects of medication treatment. To address the burden induced by ADRs, a preventive approach is necessary whereby clinicians are provided with new data-driven decision-support systems to foresee the factors leading to ADRs and plan precautionary activities effectively. We develop a multi-agent system which monitors the factors leading to the onset of ADRs using information found in the patient records in a hospital setting. The system uses a fuzzy rule-based reasoning engine utilising decision rules developed by clinicians. We evaluate the ability of the framework to identify the cause of ADRs from patient records in a case study involving records of metal health patients. Our work is the first preventive agent-based aid tool.