
****Trendy - A Kenyan Twitter App**** This is only a follow up to my previous post complete with screenshots. The objective is that I need the following info. emailed to my address: 1. If you are on Twitter, your @ address, a short description of youself (company), websites, other info. and how you'd like to be classified...refer to screenshot attached i.e entertainement, business, news..... 2. Ideas on how to filter out or classify spam messages and rate tweets based on the source or user. Source of tweet is classified as either --- spam, unknown, trusted You can suggest any other i'm missing. 3. How to improve the Suggested Users List - Currently its based on the No. of Tweets an user sends within a specified period plus the number of mentions / retweets for a user. This is an upgrade from the previous one which only checked the total tweets of a user to determine activity. 4. How to improve tagging of trending words. Someone suggested 'Predictive Tagging' a highly advanced interface. any ways of doing it on a smaller scale woud be highly appreciated. i.e "Reports of an accident along Jogoo road", I only need to filter out "accident" and "Jogoo Road". A simple algorithm to do the above would be highly appreciated. APP INFO. Trendy is a windows based application that will work mainly on xp system. Not sure about Win 7 and Vista. Earliest expected beta release date would be late next week likely after i've fixed ugs and added your suggestions.
participants (1)
-
Jacob Ayienda