Typically, there are many channels of feedback available for commuters to relay their sentiments towards certain issues related to the public transportation system in Singapore. However, the massive flow of feedback is usually difficult and tedious to analyze.
The challenges include:
The objectives of this application are:
To analyze bus transportation system through Twitter
To extract bus information (bus service, road and bus stop)
To detect events and convey information on them
To understand sentiments towards bus services
To develop a bus analytics system to profile bus services and their trends
Technology Features, Specifications and Advantages
This technology compares and implements appropriate machine learning approaches (Support Vector Machine, Logistic Regression and Conditional Random Fields) for event classification, entity extraction and state-of-the-art Stanford sentiment analysis (Stanford Sentiment Treebank and Recursive Neural Tensor Network) on unstructured Tweets. Using these technologies, Information Retrieval and Machine Learning approaches, we extract structured information from these unstructured Tweets.
This technology is useful for Public Transportation System providers and all service-oriented systems that involves automated text/information retrieval to gather micro-event feedback from commuters or consumers, respectively.
For example, it can be applied to other systems like:
Public Transportation System – Mass Rapid Transport
Extract realtime commuter feedback
Obtain realtime traffic news from commuters
Crowdsense commuter satisfaction level
In this era of massive information flow, the traditional way of conducting survey is too slow and the speed of extraction of information is crucial. Customers will be able automate the extract of information from massive unstructured Tweets in realtime. The automation involves event classification, entity extraction and sentiment detection.