The Singapore News Economic Sentiments Index (SNES), developed by the Ministry of Trade and Industry,measures economic sentiments as reflected in local newspaper articles. Specifically, it taps on daily newspaper articles to provide a high-frequency and real-time indication of economic sentiments in Singapore. This article highlights recent trends in the SNES, and explores the key drivers of these trends.
The SNES is constructed by first identifying economic-related articles in local newspapers. The news articles are then analysed at the sentence level using a proprietary polarity lexicon to determine the sentiments reflected in the sentence. Specifically, a numerical score is generated for each sentence to determine if it conveys positive (i.e., larger than 0), negative (i.e., smaller than 0) or neutral (i.e., equal to 0) sentiments. Contextual valence shifters are taken into account, to adjust for the effects of negators and amplifiers. For instance, words like “very” or “extremely” amplify the sentiment score, while words like “not” negate the score. The scores of all the sentiments-bearing sentences in an article are then averaged, and the scores for all the articles are aggregated based on the desired time frequency (e.g., daily, monthly or quarterly) to generate the corresponding SNES index.
Please click here for the full article.