The purpose of this Software Requirement Specifications (SRS) document is to provide a description about the features and use cases of the Tickets Tonight App 1.0. It will cover all functional and nonfunctional requirements of the product, as well as its UI and technologies being used. This document also details the internal software interfaces and external dependencies and interfaces. This document will describe the whole application system.
The goal of Tickets Tonight, as its name suggests, is to provide event recommendation, event search and ticket purchase service which are oriented to last minute decisions. Ticketmaster has been generally viewed as an endpoint for users to simply purchase the tickets to an event they already have in mind. Now, with this application, Ticketmaster can extend its services to even the event choosing phase and event discovery phase of a user’s whole purchasing process. Ticketmaster can better refine and cater to users who want to buy tickets at the last minute. Ticketmaster has provided sample data through xml files that contain artist, venue, category, and events data. Ticketmaster has also provided affinity data which denotes several recommended artists and their scores for a given artist. These data will be used to generate events available and provide users recommendations. Tickets Tonight will also send notifications to inform users about events which are still available tonight, based on users’ location, preference, event date and budget, and lead users to ticket purchase page once they have decided to go.
Tickets Tonight will use Parse.com to store the data provided by Ticketmaster. In order to access the data stored, Ticket Tonight will use Parses iOS SDK. To parse the given data from Ticketmaster, we will be using the python XML library and JSON library. Ticket Tonight app also uses iOS view controller, including [TabBarController](https://developer.apple.com/library/ios/documentation/WindowsViews/Conceptual/ViewControl lerCatalog/Chapters/TabBarControllers.html) and [TableView](https://developer.apple.com/library/ios/documentation/UserExperience/Conceptual/TableView_ iPhone/CreateConfigureTableView/CreateConfigureTableView.htm). In order to populate the location map on Apple Maps, we use the [CLGeocoder object](https://developer.apple.com/library/mac/documentation/CoreLocation/Reference/CLGeocoder_ class/index.html)
Tickets Tonight is a follow-on member of Ticketmaster’s mobile app for ticket sales. Ticket Tonight will be focused and catered to users making last minute decisions to go to an event. Tickets Tonight will use the event, artist, category, venue, and affinity data provided by Ticketmaster to achieve its purpose. The data given by Ticketmaster will allow the Tickets Tonight to provide the user actual events that they may be interested in. If Ticketmaster desires, this application could be integrated with their actual database instead of pre-generated data. In order to better cater to the user, the application will use the user’s favorite artists to generate more relevant recommendations of events. Ticketmaster could later link their user database to this application but without any given API, this version of Tickets Tonight will generate its own user group separate from Ticketmaster users. Tickets Tonight’s user data and favorite artists will be stored on Parse. For preliminary and testing purposes, the application will use anonymous users and will not store actual user favorites.
The overall application will have a Feed view, a Favorites view, a Explore view and a Settings(More) view. More details of the following views will be provided in Section 3. The following descriptions provide a high level summary. Each of the following views will be a view inside a UI tab bar controller object in the app. The views will be accessible through tabs on the bottom.
Feed View: In this view, the user would be able to see the events of their favorite artists and the events will be ordered by date. This view will be separated into recent, favorite and following tabs. The recent tab of the feed view will display the events that are happening near the user (if location turned on) and events will be sorted by date. The favorite tab in the feed view allows the user to view his or her favorited artists’ recent events. In addition, there is also the following tab which allows the user to easily track the events that he or she has followed. The events that the user has followed could be events he or she is going or simply the user just wants to track it. Furthermore, if the event that the user is following changes location, the user can easily track that change and get notified.
Favorites view: In this view, the user would be able to see the artists they have selected as their favorite artists. The user would be able to click on any of the artist cells and untoggle the favorite selection if he or she so desires. At the top of this view the user can also manually search for artists by their names to find the artist the user wants to add and start tracking. The search will research and refine the results as the user types in letters. For each key stroke we can further reduce the result pool and create a almost search-auto complete feature.
Explore view: In this view, the user will be provided a slide card presentation of potential events that the user may be interested in. The user can slide away the cards if the user is not interested. The events presented in this view is generated through the affinity data by first joining the lists of artists that are similar to the user’s favorite artists, then finding events performed by those related/similar artists. If the user does not have a lot of favorite artists, the slide cards could eventually be emptied since the given data from Ticketmaster is not as expansive as their actual database. In the case where the cards are all swept out, there is a refresh button on top for the user to reset.
Settings view: In this view, the user will be able to search artists by category, toggle location control and a logout button. The logout button currently does not implement actual logging out or logging in action since this application will be using anonymous users hence the user favorite data will be stored locally on the phone. There is also another tab to allow the user to access the events nearby directly listing out all events nearby.
Currently the main type of user will be potential ticket buyers. They will be notified when an event would be a good match for them based on their favorite artists and affinity data. The main user base will likely be users who intend on going to events last minute and have not made plans. These types of users are typically whimsical or perhaps are from out of town. This user class could potentially also be people who had plans cancelled and need to kill time. They would benefit from the explore view that generates events of artists relatable to their favorite artists based on the affinity data. These are our favored user class. Another user class could be event planners that are planning an event for their customer and intend on avoiding conflicts with other popular events or other similar artists’ events. In other words, these users use the application to find out openings and times that they could host their own events without coming into conflicting with other events. These users are not the primary users that Tickets Tonight want to satisfy as they will not likely be purchasing tickets from Ticketmaster.
Initially, the Tickets Tonight app will be built for Apple’s iOS 8 mobile operating system, but will be supported on devices with Apple’s iOS 7.0 and later. It will support all devices running iOS 8, including the new iPhone 6, iPhone 6 Plus, iPad Air 2, and iPad mini 3. The app will take advantage of these new display sizes and utilize the device’s location and TouchID sensors for a better user experience in receiving recommendations and completing ticket purchases. Given enough market demand, an Android version of the app will be developed. Tickets Tonight will also heavily rely on Parse to provide cloud services and database access on the mobile app. This software infrastructure may need to be changed if future versions decide to steer away from Parse and potentially Ticketmaster may want to provide its own infrastructure for database access.
One big limitation is the lack of access to the actual database of Ticketmaster. There is also no API’s to allow logging in to Ticketmaster or querying a user’s favorite artists. As such, the Tickets Tonight app will resort to creating its own user and have only the favorite artists stored for each user. This could limit Tickets Tonight’s portability in the future when it connects to the actual database instead of Parse. Using Parse, Tickets Tonight conforms to Parse’s iOS SDK infrastructure interface to store and query data. If an android app is to developed next, Parse also provides an infrastructure and Android SDK that could use the same Parse objects created for this app. Since no information has been given for the database, we also do not know how data curation occurs and how the affinity data will be generated. This limits portability as we cannot design how new information can be pipelined to update the data the app queries on Parse. Without actually connecting to Ticketmaster, the user cannot make an in app purchase and thus would be directed to the event page through their mobile device’s web browser. Another concern is security constraint involving parsing the XML file. If future data curation is continued through xml parsing through python then there is a security risk because python’s xml module is not secure against maliciously constructed data. Another constraint is that since Tickets Tonight is a mobile app and it requires accessing Parse, internet connection will be required. If the user is in a weak signal strength area then the app will have to rely on cached results that could be very outdated depending on when the app was last launched.
Ticketmaster did not provide us with an API to access their data, but instead provided us with two static data files: 1) an event info data feed formatted as an XML file and 2) an affinity feed formatted as a simple text file. Our implementation parses these data files and hosts them as Parse objects (PFObject), essentially a database (key-value pairs of JSON-compatible data), on Parse. Tickets Tonight will connect to Parse through the iOS SDK which provides access to objects on Parse through a query object (PFQuery).
Ticketmaster provided us with a 118.9MB XML file that includes a list of artists, event venues, event categories, and event information. The XML file is parsed through Python’s built in XML module and uploaded to Parse to allow the application to access the data through Parse’s IOS SDK. The data feed is parsed into four sub trees. On Parse, there is the artist subtree, venues subtree, category subtree and events subtree. A section of the data file is displayed below.
For each artist, Ticketmaster has provided us the artist id, the category to which this artists belongs, an URL to the artist’s image (low resolution), the artist name, and an URL to the artist page on Ticketmaster’s website.
For each venue, Ticketmaster has provided us the city of the venue, the state of the venue, the street of the venue, an URL to the venue’s image, the name of the venue, an URL to the venue’s page on Ticketmaster’s website and the zip code of the venue.
For each event, Ticketmaster has provided us the event id, the performer ids (artist ids), the category id of the event, the event date, event status, event time, on sale date, performance name, the URL to the Ticketmaster page, and the venue id.
The affinity data provided by Ticketmaster is a 11.1MB text file. The file contains an entry for artists. Each artist’s entry has a list of scores and corresponding artists. The artists under an entry are the recommended artists for the artist in question and each recommended artist is associated with a score ranging from 0 - 1.
The format of the file is
{ ArtistName {RecArtistID, RecArtistName, Score}* }*
“RecArtistID” can be related to the artist information in the event info data feed we have. The second value is the name of the recommended artist. The third value is the score ranging from 0-1, with higher score meaning that the recommended artist is a better match than other artists with lower scores. This file is parsed using our custom Python script and stored on Parse.
The major dependency for Tickets Tonight is its reliance on Parse. Tickets Tonight queries the data on Parse in order to generate its feed and explore views. If Parse cannot provide reliable service then the application interface with Parse will have to be changed and another form of data interface would be needed.
Tickets tonight will follow iOS human interface guideline. Tickets Tonight has two major views, Feed and Explore. The views are contained in a tab bar controller object and the user is given tabs at the bottom of the screen to change into the views. In the Feed view, the user can view events of the artists in their favorites. Each event is displayed as a cell with the image of the event on the left followed by the title and performer text in the middle of the cell. The cells are in iOS table layout. If the user clicks on the cell, then the event information view is displayed in an event view page which displays the image of the event at the top followed by the date of the event, follow toggle for the event (places the event under the following tab in Feed view), the ticket price range, and the location of the event. If the data provides the ticket URL, then by clicking on the ticket price cell, the app will start an in-app embedded browser that allows the user to buy the ticket on Ticketmaster’s website. The location of the event is followed by a cell of Apple Maps that pins the event on the map. In the artist view page the user can favorite or unfavorite the artist. The artist page displays the artist name on top followed by his or her picture and event cells of the artist. In the Explore view, the user is presented events of not only his or her favorites but also events of artists recommended by their favorites’ affinity data. The events are displayed in a seemingly flashcard form that the user can swipe away or click on. Each card displays the event image in the top center and the performer image and location at the bottom of the card. The card will also indicate who this artist is similar to. In the more tab, the user can find artists by category and also query all the events that are near his or her current location directly.
Tickets Tonight connects to Parse services in order to access the data provided by Ticketmaster. The Parse iOS SDK connects the app to Parse if the application id and client key is provided. The data from the XML and the affinity text file is parsed by python and stored on Parse as PFObjects. Each PFObject contains key-value pairs of JSON-compatible data. This data is schemaless, which means that we don’t need to specify ahead of time what keys exist on each PFObject. We simply set the key-value pairs we want, and the Parse backend will store it. The PFObject has a method to save the data to the Parse backend and the data can be viewed on the Parse app page data browser section. A sample is shown in the following image.
1. 
Tickets Tonight will query the data stored on Parse through a PFQuery object that returns the PFObject on Parse. The PFQuery can find the PFObject if given the object id which will be created as a field when the PFObject is instantiated. Tickets Tonight can save the user’s favorite artists through similar means of querying the PFObject and modifying it then saving it. Saving a PFObject to the Parse server can be done offline as well, where we call the saveEventually method on the object and the method will store the update on the device until a network connection is re-established.
The aforementioned interfaces of Parse outlines the overall infrastructure of communication and how Parse is used. More detailed documentation of each various method and relationships on query and object can be found in References. The data on Parse are used to generate the cells in the Feed view and the cards in the Explore view. For the Feed view of Tickets Tonight, we query the user’s favorite artists and then query their events and then order them by date and display in the Feed view. For the Explore view of Tickets Tonight, we query the user’s favorite artists and then query the affinity data to find recommended artists based on the user’s favorite artists. Once we have the recommended artists, we query their events and generate the cards for each of their event. When the user adds an artist to his or her favorite, the data is also saved to Parse.
In the actual application, each view, is handled by the UI Tab Bar controller object. The UI Tab Bar controller object displays a view corresponding to the tab selected at the bottom of the screen. Each view contained in the tab bar controller object will be linked by the tab. Under the UI Tab Bar controller object we have a Feed view controller, Favorite view controller, Explore view controller, and settings view controller. These view controllers are all table view controllers. Table view controllers in iOS displays table cells. We populate the tables views with Image table view cells which contain an image and text to the right of the image. In the Feed table view controller, we populate the view with Event cells which contain the image of the event and the title of the event. In Favorite table view controller we populate the view with artist cells which contain the image of the artist and the name of the artist. The event cells and artist cells lead to event views and artist views respectively. In the event views we display the image, artist, ticket URL and the location of the event. In the artist views we display the image of the artist and events the artist has. In order to populate the cells with actual data, we query the data on Parse through the aforementioned Parse interface. By querying Parse with the PFQuery object we can obtain the data we need in order to populate the cells. For detailed description of the object methods, the documentation for view controller objects for iOS is included in the reference section. In order to display the map in the events page, we convert the address to a geopoint using the CLGeocoder object and using it we could mark the venue on to the map. In order to open the URL inside the app, we create a webview to display it. In addition, when we query the data needed for the explore view, we first obtain the favorite artists objects then flatten them to names in order to query the affinity data, then acquire recommended artists objects, then flatten those objects into int ID’s, convert those ID’s into strings, then query the events and finally exclude the events that the user are already following.
For the data given to us by Ticketmaster, we use Python to parse them into essentially a JSON format to upload to Parse into Parse objects. For the XML we use the Python library to parse the XML into a tree like objects and create csv objects. For the text file containing affinity data, which was almost a csv file, we also use the python library to parse it into JSON formatted objects. These objects are then uploaded to Parse. Parse accepts csv and JSON to populate Parse objects. More detailed description for each of the python library methods can be found in the reference section. We parse the XML into a csv but the affinity data into JSON because for the affinity data, an artist can have an array of recommended artists and each artist has a different length of array. CSV requires column number to be the same for each entry. We could have also put all the recommended artist into one long string in a column for the CSV however, we chose JSON in order to gain easier access and without having to parse a string.
Our initial design for the project was under the assumption that we would be receiving access to Ticketmaster’s database and API. Under this assumption, we wanted to create a great overall experience that would allow the user to actually be able to purchase the ticket through the app and connect to their Ticketmaster account in order to synchronize the product line. Our original design was using node.js as our server in order to retrieve data from Ticketmaster and have node.js serve as our platform for the data we needed on the app. We also originally wanted to connect the user favorites data to the user account on Ticketmaster’s website so ended up using an anonymous user and store favorites locally.
One difficulty we had was the designing of the whole application. We really wanted an intuitive way to be able to present the data to the users. Another difficulty we had was combining all the data into a coherent application. Initially we thought about combining all the data to formulate the best recommendation for the user. Another difficulty was converting an actual address to a geopoint in order to mark the location on the map. We needed the geopoint of the location instead of addresses and currently we do this conversion on the mobile device. We could potentially convert all the event locations and save their geopoint instead of the address string.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent