Develop a use case diagram for a Twitter analytics system that we shall call Tweetanalytics. Twitter today has become a voice for many consumers, and companies and brands are very interested in learning about what their customers have to say, and also interacting with them through this medium. Furthermore several brand communities and conversations organically form around brand hashtags serving as a potential treasure-trove of information for companies. Twitter data is not only useful for merchants, retailers and brands, but there are several other applications. For example, StockTwits and PsychSignals extract tweets on financial data (stock tickers) and create indices for investor sentiment which might be useful for predicting how markets will perform. You could also develop any application on top of Twitter that you can think of!

This system will obtain information from Twitter using the TwitterAPI. API is Application Programming Interface and it is a software interface between Twitter and any other outside system that wishes to extract or use Twitters internal data. In other words, Twitter provides an API which are a set of pre-determined software commands - that Tweetanalytics will use in order to make data requests to Twitter. In turn, Twitter API recognizes the command and provides an appropriate handshake which returns some data (tweets) to Tweetanalytics. The system can perform a variety of analyses using the tweets.

For this, create a UML Use Case diagram from the following (bulletized) requirements. Every black bullet point is a main goal/task of the user should be a primary use case.

A data analyst can use Tweetanalytics to

Open a connection to the Twitter API. Whenever a connection is trying to be established, send user credentials to the Twitter API.

Send http requests for tweets to the Twitter API. This causes the Twitter API to respond back with a collection of tweets (typically sent in JSON format used for data transfer on the web).

The system should then receive the tweets:

Parse the JSON data (received from Twitter). Once the parsing is completed, the system should save the parsed fields to a database (such as MySQL that contains a Twitter table).

Perform text analyses on the data. To do so, the data has to be loaded from the database, but only when it is not already in the memory. Additionally, the analyst may choose to perform one or more of these specialized text analyses[1]:

extract annotations using Named Entity recognition,

extract geographic data using Geocoding,

extract significant phrases

extract sentiments

All of the above special analyses tasks have some shared steps related to preparing the data for special calculation (hint: best described in a parent use case).

Produce managerial reports from the analysis. Always make sure to save the files.

A manager can:

View managerial reports

Make edits to managerial reports. Every time edits are made, save the report.

Tip: You do not need to understand what each of these mean in order to try your hand at a use case diagram. Here is a summary. Named entity recognition tries to identify the following in each tweet: organization, person name, date, money, time, percent. Geocoding attempts to get the place/location of tweets that dont contain it already. Significant phrases depend on the application merchants and companies look for product/brand names in the tweet, whereas financial firms and investors look for stock ticker symbols. Finally sentiment calculators determine if what was tweeted was positive or negative. Remember, even though some formats (such as http, json, MySQL) are mentioned here, there is no need to use technical details on a use case document. The purpose of the use case diagram is to help understand user requirements what would the user like to achieve (these are primary use cases connected directly to the actor). If the system need to do additional steps to satisfy actors primary goals, then indicate that with secondary use cases. If there is an external system, depict that as an actor. Associations are solid lines, and includes and extends are dotted lines.