Jarvis+, a decentralized conversation-as-a-service platform, has published the 21st edition of its weekly report series. The latest installment covers the team’s progress in research and development as well as an update of its ongoing operations.
For development, the team worked on a search engine spider to collate and analyse information posted online. Work began on the spider in preparation to co-host the Natural Language Generation (NLG) Challenge, in conjunction with Microsoft.
Registration for the event opened earlier this month and will close on February 20th, 2019.
NLG is the inverse of Natural Language Processing (NLP), which is the process of analyzing text and extracting data. With NLG, data is automatically transformed into a written narrative, which can then be used for the purposes of business intelligence, IoT device reporting, and many other applications.
So far, it’s claimed that over 100 million related news items and comments have been collected for use in NLG. The data will reportedly be housed by the platform and used for generating data and validation sets. The spidered content also reportedly went through its first round of preliminary testing.
Other work included bug testing and documentation of its candy system, auto bot creation system, and virtual KOL. More work was completed for the platform’s virtual dialog training based on its “CNN” model.
The platform’s group quiz game is ongoing. The game allows users to answer questions in return for a chance to win Jarvis+ tokens (JAR).
To participate in the quiz, users sign up and play the game in the platform’s dedicated Telegram channel.
Jarvis+’s daily check-in game is also still available to play. Based on similar rules to the group quiz game, the daily check-in rewards users for answering a simple question every 24 hours, typically on subjects related to the Jarvis+ ecosystem, or about the platform itself.
The daily check-in game is based on a points system with points awarded based on the accuracy of the user’s answers. Each correct answer awards the user three points, while incorrect answers award only one point; points are redeemable for JAR.
Finally, users can play a fantasy soccer UEFA game with Jarvis+. The game allows users to bid on the outcome of matches with JAR with winners taking a proportionate amount of the opposing bidder’s winnings.