You partied hard on weekend till late night and on Monday morning you are rushing to reach the office on time, how many activities you compromise with just for the sake of reaching office on time? breakfast? shower? or reading the news! There are many people like me for whom every morning is rush hour irrespective of me partying hard previous night or not and we miss reading news on daily basis. We try to read a thick news paper or news app while commuting. But how do you decide if the selected news is relevant and interesting to you or not? by reading the headline, by looking at images and by reading first few lines of the news. If it is relevant then you read it if it isn't you go back to the main menu, select another news and repeat the process. What if someone gives you a brief summary of full-fledged news and then you can decide if you want to read it or not? That is what City Varta is, aggregating and summarizing customized news just for you.
People get busy in their day today life. When you get busier, you get less time to read. Whenever a small amount of factual information gets fed to the brain, we feel good about it, we feel more intelligent. This is why pages like What The Fuck Facts or 10 things you didn’t know about X, are so addictive and are not only click baits. People like the feeling of being intelligent (who wouldn’t). But most of the facts are useless information, for example: “A Mosquito has 47 teeth”, this is a pretty cool fact to know, but it’s totally useless. What if we combine facts with knowledge that would be useful in day today life? Keeping news short, and delivering them as facts. This way factual information becomes knowledge and usable in day to day life. City Varta in a way delivers news in the way they look short and cool without twisting the actual news. I talked and interviewed people from my workspace and other who have morning rush to gather their goals and pain points about their current news delivery system. Some didn't have time to read the news at all, and some didn't want to read an irrelevant news, many people wanted news content relevant to their city. Moreover, our city (Aurangabad) has no news application which could deliver city news.
Following were the findings of my user research.
Balsamiq mockups before finalizing the design and starting development.
Rather than creating a prototype in Axure or InVision, I thought to create a working native android application and get it out in Market (in the cheapest way possible!). The app was small, so I built the server database in MySQL coded rest APIs and admin panel in PHP with the help of the slim framework. At first, I thought I should write server code in python, as implementation would be much faster, but I dropped the idea because hosting was not cheap. The mobile end is on Android, where I used SQLite to store the news. I wrote a simple algorithm, where every time app connects to the server only the latest news which is added on the server will be fetched and stored in the app’s database, and not the older ones. Looking at the uses cases, most of the users were going to read on the run, and mobile internet connection is not that strong in the Aurangabad. So app needed to have the capability to give summaries of news when internet connection was not present. So, SQLite database was essential, as using that I was able to deliver offline news to the users.
How do I add news? basically, I check the news in newspapers, or at other sources, I summarize it in 60-80 words, and then I add the data to the server, and the mobile app pulls it in JSON format.
For Android, I used many third party libraries like Picasso, volley for networking, input SDP, etc.
Here's the github link of the code.
At first, everyone thought I wrote some ML or NLP algorithm which summarizes the paragraphs of a news article in 80 words, but when I told them I am manually summarizing it, their next question was how will you scale it? I asked them, 'Have you heard of “New York Times” or "The Wall Street Journals"?' they have an army of editors who does it for them, and they are doing just fine. I am still figuring out the revenue generation part, customized ads maybe, but will not go for android ads, they are so badly designed.