Abstract:
The massive volume of Twitter data has attracted much attention of researchers to study their correlation with stock market. Tweets with stock symbols can be identified b...Show MoreMetadata
Abstract:
The massive volume of Twitter data has attracted much attention of researchers to study their correlation with stock market. Tweets with stock symbols can be identified by the prefix with dollar sign or by using some complex techniques. In this paper, we focus on discovering NASDAQ stock symbols in a stream of tweets. We propose a simple but effective methodology to recognize the stock symbols. Stock symbols from NASDAQ company list, WordNet, Wikipedia and sample tweets, as well as a classic method of collocation discovery are employed to filter stock-related tweets. Experimental evaluations show that our methodology outperforms the baseline approach for recognizing NASDAQ stock symbols.
Date of Conference: 30-31 January 2014
Date Added to IEEE Xplore: 24 March 2014
ISBN Information: