Original Content Generation for Assistants, Collaboration

Can content write itself in a useful way?  Note this neural method is research,  technical and apparently incomplete. Yet still points to some useful directions.  Ultimately this could lead to a method that would answer questions from a corpus of information, ultimately what you want to get from an assistant,  or as an agent that is working in collaboration with humans to answer support questions.   Imagine such a method using a collection of FAQs or 'solved'  inquiries to create new solutions.  This might also be a way to create 'Content Based Reasoning' solutions based with new and old resources.

Google’s New Algorithm Creates Original Articles From Your Content  in Search Engine Journal

Google has published research of a new algorithm that can take yours and your competitor’s webpages and generate “coherent” articles. By creating original content, Google’s new algorithm can answer a user’s question without having to send them to another webpage.

How Does the Paraphrasing Algo Work?

Google’s new algorithm works by summarizing web content using an algorithm that “extracts” your content then tosses out the irrelevant parts. This is similar to the algorithms used to generate featured snippets.

These are called “extractive summaries” because they extract content from webpages. Extractive Summaries are like a reduction of the original text to the most important sentences.

Afterwards, this algorithm then uses another kind of algorithm called an Abstractive Summary. Abstractive summaries are a form of paraphrasing.

A downside of artificial paraphrasing (abstractive summaries) is that almost a third of the summaries contain fake facts.

Here is more information about extractive summaries, Faithful to the Original: Fact Aware Neural Abstractive Summarization

Google’s new research has discovered a way to join the best of both approaches. They use “extractive summaries” to extract the important facts from web documents and then apply the “abstractive” approach to paraphrase the content. This approach creates a new document based on the information found on the web, creating Google’s own version of Wikipedia.

Google’s new algorithm is described in a research paper titled, Generating Wikipedia by Summarizing Long Sequences  ... "    

Case Based Reasoning Driving Chatbots

I was reminded about the possibility of using, even powering chatbots using case based reasoning (CBR) methods and intelligence architecture.  We had looked at this for some simple cases by constructing decision trees and building on those structures.  This is also of value because CBR data is easier to maintain and update with new learning.    Earlier post on this topic.  Anyone done something related?

Case Based Reasoning and Chatbots

 Resources for integrating Case based reasoning (CBR) and bots. The integration could be a natural way to deliver intelligence in a bot.  We used CBR in a number of early applications.

 

" ... The main goal of jCOLIBRI is to provide a reference platform for developing CBR applications. With the term “platform” we mean that it defines a clear architecture to design CBR systems plus a reference implementation of the components required to build them.  This architecture has been designed to be extensible and reusable across different domains and CBR families.

''... Although the main scope of the framework is the research community, a relevant feature of jCOLIBRI is its support for large scale commercial applications. It means that our software can be used not only to do rapid prototyping of CBR systems but also for developing applications that will be deployed in real scenarios. This feature is exemplified, for example, by its straightforward integration into web environments.

jCOLIBRI takes advantage of several years of experience from the first version released in 2005. Therefore, the current architecture solves many problems found in its predecessor but reuses the design choices that demonstrated to be good ideas. ..  ' 

Via Jim Spohrer, some examples of CBR architectures:

jCOLIBIRI (CBR)   http://gaia.fdi.ucm.es/research/colibri/jcolibri

MyCBR (CBR) http://www.mycbr-project.net

The image at the upper right is in an article from the University of Trier.

Following Watson Facing Cybercrime

Continuing to follow Watson to determine what the ideal application is.  Is it mostly about big and volatile knowledge, appropriately indexed?   Leveraged with machine learning.  To find appropriate cases for reapplication?  That's what I have seen so far as part of application proposals.   Not AI ... but certainly a form of 'Practical Intelligence'.    And maintaining that knowledge via focused learning.

Example, Recently in  Wired:

" ... IBM announced that Watson is taking its cognitive learning chops to the cloud, where it’ll apply them to analyzing, identifying, and (hopefully) preventing cybersecurity threats. But first, it’s going to have to learn. Fast. .... 

There are already plenty of computer-enhanced approaches to combating cybercrime, most of which involve identifying outliers or abnormalities—like when a user logs a few too many failed password attempts—and determining whether those constitute some sort of threat.

Collecting and analyzing this type of data can and does work. It’s not ideal, though. First, there’s simply too much of it; according to a recent IBM report, the average organization sees over 200,000 pieces of security event data every single day. There’s simply no way to keep up with it all. And while solutions like MIT’s recent AI2 can trim down the number of incidents a human researcher needs to sift through, there’s still the fact that the data points being considered are only a small part of the picture.  ... " 

Hyper Personalization

Interesting startup direction ...  Reminds me of case based reasoning approaches ....

" .... Analytics passed a major milestone on the road to ubiquity last week with the launch of new capabilities that extend the reach of data-driven decisions even deeper into our digital lives. An up-and-coming startup called Jivox Inc. set the ball rolling with the introduction of an advertising platform that promises to provide the ultimate customized experience.

Dynamic Canvas,  as the new system is called, assembles pre-supplied graphics and copy into one of over 3,000 potential promotional formats based on the specific preferences of each viewer. Jivox gleans that information from a combination of metrics ranging from basic details such as age and gender to much more precise data points like local weather to produce what is touted as a much more through understanding of buyer expectations.   ....  "

More Tales of Chef Watson

Have spoken to several people recently on the now free and public cognitive demonstration system.   All the comments I have seen, by non professional cooks, have been positive.  Now here another use case.  Includes a recipe:  Chef Watson's Spicy Avocado Brussels Sprouts.    A book on the system is coming out soon.  A classic example of a domain focused, constraint-based advisory system, based on stored cases.  A form of Case Based Reasoning (CBR).

Mining Massive Databases

 Vincent Granville started the discussion "Mining Massive Datasets - Coursera Course (Stanford)" in the group Analytics Courses on AnalyticBridge.     Now looking at the scale of databases more closely.

Bad Data Handbook

I like the book title premise. Having just been involved in a project that included dealing with bad data, however it is defined.    Also to be considered:  Data and metadata that is not or miss-connected to data in use. Or stated as  "From cranky storage to poor representation to misguided policy, there are many paths to bad data. Bottom line? Bad data is data that gets in the way. This book explains effective ways to get around it. ...  ".   Don't have the book, so can't say it delivers.   More about the book in Analyticbridge.