No abstract

Twitter is the most famous on-line microblogging service now. People can post (tweet) what they are doing in 140 characters. Since Twitter posts (tweets) reflect what people are looking, hearing, feeling and so on, we can obtain information about Real-world phenomena through the large amount of tweets. In other words, Twitter can be regarded as a sensor of Real-world phenomena including natural phenomena such as hay fever. This motivated us to investigate whether can Twitter be an alternative of Real-world Sensor. In this paper, we first describe about our system which collects and analyzes tweets in order to generates a hay fever map just like as a weather report map. There are some difficulties such as location estimation and normalization of number of tweets. Using the output of the system, we discuss the comparison with actual pollen data gathered by real sensors. The result shows that Twitter can reflect natural phenomena in some particular areas.

No abstract

No abstract