… so then we thought: “Why don’t we just make a robot that reads all of this news for us?” So then we wrote some software & thought some more …
There have been several enhancements within the Computer Science arena over the last ten years. This is due, in part, to the availability of exponentially faster processing power and, as well, the creation of software libraries and enhanced algorithms.
Some identified technologies include the following:
- Natural Language Processing
- Machine Learning using Tensor (Flow) Networks.
- Greater Efficiency of Least-Cost Locality & other Traveling Salesman class of problems.
- Quantum Computing… Well, maybe one day soon.
These powerful technologies can be combined in new and meaningful ways. We’re working on these technologies here.
With over 90 LI News Sources now get sorted, and tagged, automatically based upon previous news content, identified keywords, and short phrases. A one stop shop where fewer clicks allow more tricks.
We do anticipate many challenges and achieve many milestones while we refine & develop this project further. The following line items are highlights.
- Identify existing Long Island Technology New Sources and store them for processing. *
- Contextually categorize individual stories automatically by using a basic form of Natural Language Processing. *
- Identify specific keywords and terms that are important to the intended audience. Take these keywords and terms to tag individual news items. *
- Trend & behavioral analysis with news stories and user interactions with them. This will provide meaningful statistics to increase the value of the User Experience and provide the highest quality content heuristically.*
- Facial & Object recognition on Images processed by the LITechNews system. News Articles are now using meta tagging by Detected Object (Ex: Automated Teller Machine, Demonstration or Pedestrian Crossing) or by Facial Recognition (Ex: Al Gore, Donald Trump, Tim Cook or Vladamir Putin). This software has analyzed thousands of images thus far and we will keep integrating this new layer of meta-data.
…. and a few other ideas that currently live on a white board.
Contact the Team here for additional information of if you believe you can contribute.
Some of these Technical Milestones are already implemented and are being analyzed, updated and tested here. Other identified milestones are being researched & tested so that additional features can, as well, be field tested here.
We encourage you to help us. Here are a few examples how you can do that.
- Liking & Following LITechNews & our sister twitter account LITechNewsPosts.
- Tell your colleagues, friends, loved ones & fellow nerds about what we’re doing here!
- If you have technical experience with technologies such as OpenAI, TensorFlow, Keras, Theano or are interested in Machine/Deep Learning systems then we want to speak with you.
LI Tech News Statistics
LI Tech News Launch Date: October 7th, 2016
Total News Articles Processed: 142661
Total Article Summaries: 20996
Total Articles Analyzed by NLP: 115141
Total Images Facially Recognized: 15449
Total Images Visually Itemized: 59016
Total Posts Analyzed for Geographic Data (GIS Extraction): 141874
Average Posts Per Day: 351
LITechNews Fact #1: The LI Tech News robotic engine fetches updated news stories every 15 minutes, 672 times a week.