| المستخلص: |
Recently, there is a service which indeed associated with Web2.0 ; this service is micro-blog. They have their features which make them different from traditional blogs. They allow their users to write and read short massages and share these messages with other users. The most popular example of micro-blogs is Twitter. There are many characteristics that micro-blogs have which make them distinguishing from other social media such as blogs or even from the traditional web pages. One of these characteristics is the real-time nature of micro-blogs which make them a good and rich source of news and hot topic. Another characteristic is the size of the content in micro-blogs which is one of the challenges that facing micro-blogs. This research will investigate that whether or not using the Twitter’s features will improve the returned result. The features that will be covered in this research are hashtag feature, URL feature, user name tag feature, inverse document frequency (IDF) feature, tweet language feature, tweet length feature and user tweet number feature. In order to find answers to the research question, we conduct the experiment that divided into two main phases. The first phase is to examine each feature individually in order to investigate their impact on the returned results as well as to obtain the ideal weights for each one of these features. The second phase is to deal with these features as a one combination instead of dealing with them independently; therefore, we combine all these features in one formula. Using the traditional ranking approaches alone with micro-blogs might not be the ideal option. Therefore, one of this research aims is to help the researchers who concern about ranking results based on a given query or flirting and clustering data on micro-blogs environment to think about creating new ranking strategies.
|
