space. There are startups, companies and VCs who are recently exploring the Indian potential and creativity. • However, the news space in India is fairly neglected in this race of digital innovation and creativity. • The major players are old, established news agencies - most of them dump the articles and stories from their channels and newspapers to an app for online viewership. There’s not much technical advancement, analytics and positive UX there. • The new players are startups which are mostly aligned with the goal of micro-blogging and news summary. Although a neat idea, there are some shortcomings to it. Also, most of the processes are manual which pose issues of scalability, human bias and human error.
ff ort Not a lot of technical advancement which limits coverage. Sub par UX The apps do not have a rich user experience. Poor search Search is really slow, often incorrect and not context intuitive. No personalisation or recommendation A user is subjected to all news. Some apps create recommendations but those require the user to manually mark their areas of interest with each article instead of inferring it. No scalability It could take hours for a latest news item to appear on the app. Similarly, not all realms of news are covered since the entire process is mostly manual. No correlation analysis Currently, no news platform provides impact analytics or relation uncovering between news pieces.
platform by addressing the shortcomings of current market products by automating processes and using intelligent systems. • This would be a technology start-up and not a news agency i.e. we would not be producing news but provide a platform for news collection and access.
of the app and will be rolled out in subsequent releases. The sub-points give a brief idea of how they will be achieved: • Fast & relevant search • A semantic search platform to understand queries based on keywords, context, trending etc. • Elasticsearch (Lucene) based inverted index for faster retrieval • Recommendation and personalization • Apart from top news, only relevant articles based on ontologies will be shown. Eg: If you search about Indo-China, you’re most likely to be interested in recent news of India-China-Nepal politics, China-Bhutan border, USA-China trade deal. • Infer user interests from KPIs like search queries, clicks, shares, read time etc. • Fast latest news • News collection from varied sources - well-regarded news websites, newspapers, twitter feeds, direct govt notices and statements, state and district level news, press releases etc. • Due to automated operations and diverse news sources - the latest news would reach users in real time.
distributed news items. Eg: Impact of a news on fi nancial markets. • Resolve ambiguity between news items. • Sentiment analysis. • Summarizing information. • Better segment classi fi cation • A deeper taxonomy for news classi fi cation for targeting user interests more precisely. • Fake news obstruction • Use AI to combat fake news by building intelligent system which will score webpages, rank news sources, use keyword analytics to identify fake items etc. • Privacy • Anonymised user analytics • Security • Encrypted user information • Secure data storage • Transport layer security(SSL)
deal with issues of creativity, ethics, analytics etc. But the mundane and error prone aspects of news would be handled by the AI systems like: • Collection of news from disparate sources (Crawlers, scrapers) • Making sense of the news, extracting keywords (Parsers, NLP) • Storing news securely (Scalable storage) • Categorising news (Classi fi cation) • Matching di ff erent information, de-deduplicating news • Finding relationships between news (Ontologies) • Extracting references to other documents (Ontologies, crawlers) • Fake news detection • Build summary, derive statistics • It will also solve issues of scalability by allowing to cover more news stories from wider segments and locations. • The search would be more relevant and faster. • Each user will have a personalised news feed. Technology benefits
today. • People and businesses are looking for relevant and reliable information as soon as possible. • This product checks everything missing in the market today and can really transform data analytics and information dissemination.
• B2B • Use the rich Knowledge Graph as a repository of information and sell category speci fi c data to businesses/individuals/institutions. Eg: Finance/trading related information as a service for stock market individuals/ fi rms, Govt tender/business taxes information to companies. • Reliable news is a big deal today. Companies like Facebook, Google are constantly looking for fi ghting fake news on their platforms by collaborating with companies in local markets. This clean and rich ontology-based data repository could be a big sell to them.
- In Progress • Est Date of Completion: Aug end • Search system (Lucene based) • Est Date of Completion: Sept end • Recommendation engine • Est Date of Completion: Mid Nov • User interface / Mobile application • Can happen parallelly. Looking for interns.
ayushi0430
csc
babylonzoo
r5ni4
travelhupsupport
cmbtech
tonotej274
ballas_inc
quontier
jictech_saiyo
nepap
onecareer
kt0118
marcelosomers
destraynor
csswizardry
trallard
matthewcrist
smashingmag
carmenintech
philnash
bkeepers
keithpitt
marktimemedia
mclloyd
