![]() Our approach innovates by using the network of hashtag co-occurrence to discover all the hashtags relevant to the elections and to assess the consistency of our hashtag classification. We develop a supervised learning approach to measure the opinion of Twitter users where we do not try to classify tweets as expressing positive or negative sentiment, but as supporting or opposing one of two top candidates: Hillary Clinton and Donald Trump. ![]() For this purpose, the setting of the 2016 US Presidential Election allows us to have access to a large aggregate of traditional surveys performed at regular intervals to compare with our Twitter opinion trend. Here, we are interested to capture the opinion trend in Twitter and compare it to opinion time series from independent off-line surveys. Despite all these improvements, opinion time series derived from Twitter have not been validated so far with any traditional polling performed at the large scale. Beauchamp 35 extracted significant textual features from Twitter by fitting a model to existing polls and showed that these features improved state level polls prediction. 34, 37, 38 used a supervised machine learning approach based on a hand labeled training set to estimate the proportion of tweets in favor of each candidate in the 2012 US election and the 2012 Italian center-left primaries. Moreover, evidences suggest that it is possible to differentiate Republican and Democrat Twitter users based only on their usage of words 36. Recent works 34, 35 have shown that by going beyond sentiment analysis, and by considering all the terms used in tweets, even the terms usually considered neutral, a more accurate measurement of the Twitter opinion during the 2012 US election was possible. In this case, correctly capturing the context of the events is crucial to measure support. However, not only does lexicon-based approach perform poorly on the informal, unstructured, sometimes ironic, language of Twitter 33, but classifying sentiment as positive or negative does not allow one to differentiate simple attention from political support, especially during political scandals 29. ![]() Lexicon-based sentiment analysis 30, 31, 32 has also been used to improve this approach by attributing a positive or negative sentiment to the tweets containing mentions of the candidates or parties. Indeed, most work compare the volume of tweets, or mentions, related to the different candidates with traditional polls or election results. One of the main criticisms is that instead of measuring the political support for a candidate or a party, they measure the political attention toward it, those two concepts being not necessarily correlated. However, these initial investigations achieved only mixed results when compared to traditional surveys and engendered a number of critical studies 20, 28, 29 questioning their methods and findings. With the increasing importance of Twitter in political discussions, a considerable number of studies 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27 also investigated the possibility to analyze political processes and predict political elections from data collected on Twitter. Several works have shown the potential of online social media, in particular of the microblogging platform Twitter, for analyzing the public sentiment in general 1, 2, 3, 4 or to predict stock markets movements or sales performance 5, 6, 7, 8, 9. Our analytics unleash the power of Twitter to uncover social trends from elections, brands to political movements, and at a fraction of the cost of traditional surveys. We investigate the dynamics of the social network formed by the interactions among millions of Twitter supporters and infer the support of each user to the presidential candidates. The Twitter opinion trend follows the aggregated NYT polls with remarkable accuracy. We validate our method in the context of 2016 US Presidential Election by comparing the Twitter opinion trend with the New York Times National Polling Average, representing an aggregate of hundreds of independent traditional polls. Here we develop a method to infer the opinion of Twitter users by using a combination of statistical physics of complex networks and machine learning based on hashtags co-occurrence to build an in-domain training set of the order of a million tweets. Despite the large amount of work addressing this question, there has been no clear validation of online social media opinion trend with traditional surveys. Measuring and forecasting opinion trends from real-time social media is a long-standing goal of big-data analytics.
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |