The Identification of the Top-N Most Frequent @mentions and #hashtags in the 20 million Turkish Tweets
In this homework, we are going to indentify the top-N most frequent @mention and #hashtag entities. The dataset contains 20 million Turkish Tweets and can be downloded from here.
Please read the write-up: What are @mentions and #hashtags?
Your project must be a valid maven project. mvn clean package must produce an executable jar file named trending.jar under the target directory.
This can be done via maven plugins such as shade or assembly plugin.
Following command line options must be supported.
| Option | Description |
|---|---|
| -n, --number | The number of entities to display. [defaults to 10] |
| -e, --entity | The name of the entity (e.g., hashtag, mention or emoji). [defaults to hashtag] |
| -r, --reverse | Reverse the comparison (e.g., display most infrequent entities). |
| -i, --ignore-case | Fold upper case to lower case characters (e.g., collate #AnadoluÜniversitesi and #anadoluÜniversitesi). |
The result will be printed to the standard output in the format of two columns (entity \t frequency) separated by a tab.
For example, java -jar target/trending.jar -n 20 -e mention -i Tweets.txt will display top-20 mentions in decreasing order by their frequency.
Another example, java -jar target/trending.jar -r Tweets.txt will display 10 hashtags in increasing order by their frequency.
P.S. To parse command line arguments, you can use JewelCLI library.
P.P.S: Optional parameter finalName can be used to change the name of the shaded artifactId.
Hint: Regular Expressions can be used for detecting entities.
❗ If you don't follow the aforementioned conventions, you will receive very poor grades (even if you think that your code works perfectly).