MarkWeber 100000KTXK 516 ViewsModified on by MarkWeber
MarkWeber 100000KTXK 424 Views
MarkWeber 100000KTXK 853 Views
Developers new to working with Watson APIs and the Watson Developer Cloud can use the starter kits, sample applications and technical demos to start to learn how wo use developer resources available and the Watson APIs.
The Watson Developer Cloud website is: https://www.ibm.com/watson/developercloud/
• At top, click on Starter Kits - these are sample apps using specific Watson cognitive services. Click on the Text Message Chatbot app – https://text-bot.mybluemix.net - it has combined multiple services in a single solution, using the Conversation service and the AlchemyLanguage service. It also uses external data sources for weather data.
▪ You can access complete source code for the example app and in GitHub at https://github.com/watson-developer-cloud/text-bot - and get an overview of the solution architecture including use of extension points for bots with weather services, external data sources and messaging systems such as Facebook Messenger as well as Twilio. This is in the best practices section at https://github.com/watson-developer-cloud/text-bot#best-practices, including a guide to installing and debugging the application. Note that the application gets access to the Watson services instance running in Bluemix by using service access credentials accessed thru environment variables.
The Watson Developer Cloud site Developer Tools link provides access to complete documentation for each of the cognitive services, API reference and language specific SDKs that are available, as well as Community and Support options, expert forums, slack live chat, and additional resources
Watson cognitive services and APIs are available as follows:
To see one of the natural language services working, you can use the Natural Language Understanding demo, you can use an example of a document, webpage, or an article URL and the demo app will use the NLU service to show entities, emotion, and concepts in the document content. The demo app also shows the JSON payloads output from the API service calls. For most interesting examples, it is recommended to use content associated with a person and their written or spoken/transcribed conversation, such as an interview. The NLU service can be used for use cases are with social media to analyze sentiment regarding a brand, company or products, and news articles for example.
Now lets look at an app that has integrated the Watson conversation service with some other services to provide an example of a more realistic application than some of the starter kits or demo applications. As an example, Ana Insurance Chatbot at ana-insurance-bot – http://vimeo.com/196771046 click here to try it yourself. – http://cloudlife.mybluemix.net This is a node.js app that uses the Watson conversation service with the SDK for Node integrated with Mongo and Passport for external data sources and authentication services. Type in tell me about my benefits, orthodontics, I just had a filling, enter amount, doctor and date to submit claim
You can build the conversation service into an app using the APIs provided (REST) or via available SDK’s natively (Java, Node, etc). Applications can be built in Bluemix on any of the available runtimes using either REST API’s or the provided SDKs for Watson services. Apps do not need to be hosted on Bluemix to use Watson services, they can do so from any platform or cloud.
You instantiate an endpoint for the Watson service instance in your app with the service endpoint URL and credentials using the API’s provided, this sets up the connection with this service instance and it can provide your app with user input data, meta-data about the dialog, context data and Watson provided data related to the dialog. To see the JSON payload data returned by the conversation service, you can use the simple conversation demo app at https://conversation-simple.mybluemix.net note that the API’s being used are shown on the right side – JSON payloads including user input data, the step in the dialog flow, the intent that was matched, confidence level, entities, nodes visited, and context variables.
You can combine additional Watson services along with the conversation service to provide more capabilities, for example, Speech to Text could be used to provide a voice interface for conversation in addition to other channels such as email, SMS, Facebook Messenger and others. As another example, the Tone Analyzer service could be used to detect angry customers that need to be routed to a live agent.
Lets look at another Watson cognitive service, the Discovery service - this example application uses both Discovery and Conversation services - https://conversation-with-discovery-within-ui.mybluemix.net/ You can use the conversational interface to ask the application to play music, turn on the lights, adjust the heat and AC and other tasks. On the right hand side, the matched intents & entities detected are shown. When you ask for assistance with something that is outside the scope of the predefined dialog nodes in the conversation, for instance such as “how can I check tire pressure” – the application invokes the Discovery service APIs to retrieve relevant documents. Make note of the results JSON payloads on the right side.
Use cases for Retrieving information using the Watson Discovery service or Retrieve and Rank can include to provide a very effective self-service app for employees or B2B customers needing to perform complex tasks such as vehicle or aircraft maintenance and repair, integrating service manuals and videos. https://conversation-enhanced.mybluemix.net/
In a related use case, other services such as Retrieve and Rank with Watson Machine Learning for predictive analytics and the Watson Tradeoff Analysis service can be used in combination for a personalized product recommendations use case where recommendations could be delivered thru multiple channels such as email, SMS, Facebook Messenger, chat, and voice. In fact Wayblazer, a startup company founded by the leaders of Kayak.com, is using Watson services to provide personalized travel recommendations for hotels, flights, destinations and vacations as a personal travel concierge.
Simmons-Scott 110000DMJ3 2,800 Views
First - let me say that Stephen Wolfram's blog on Watson is old .. but someone mentioned it to me when we were discussing the potential for Watson Systems ... for those of you who have not read it - it is at http://blog.stephenwolfram.com/2011/01/jeopardy-ibm-and-wolframalpha/ ... I came away from it with a real bad taste in my mouth.
As background - my initial graduate work in college was as a biochem/genetics researcher in my Master degree -- I did some of the early work on alpha-lactalbumin and potential for early diagnosis of cancer (http://en.wikipedia.org/wiki/Alpha-lactalbumin) -- the reason I left research (then joined the Air Force and eventually ended up here in IBM) was that I thought the research community was an environment with TMP ... Too Many Primadonnas. Doctoral candidates indirectly "sniping" at others research and putting it down in an attempt to raise their profile ... That said - I have many good friends and colleagues in IBM Research and I do not think they personify these characteristics ... for the most part..... ;-)
So - my thread is that the above blog post by Mr Wolfram comes off as both petty and pretty childish ... it suffers from the short sighted primadonna approach I saw in genetics research - that said, it was done over 2 years ago - maybe Mr Wolfram has gotten enlightened -- not sure ...
In his blog - Mr Wolfram fails to understand the key technologies we are bringing together here -- the semantic parsing in our implementation of deep NLP aspects of Watson are not being considered nor do I see a discussion on the aspects of Content Search, Big Data and Analytics -- though Mr Wolfram seems to praise Watson - it seems to be more positioned to reinforce his view that he has a "better mousetrap" and to ride our coattails with indirect referencing of our work ... and slighted comments that evidence a lack of understanding of the richness of our approach
Mr Wolfram -- in short - Watson is MUCH more that a search engine (which seems to be your premise) -- in his post, he even does a comparison with Google and Bing. Bottom line - I am not sure on why he is so dismissive -- this may be a dated perspective -- I certainly hope so.
IBM Research is building a radically different way to look at computing ... this is not your fathers AI solution -- this is the advent of Cognitive Computing ... and hold on to your seat -- it promises to be an exciting ride ... and those of us leading this charge (like IBM) will be amazed by the business applications we will see based on the Watson technologies as well as the changes in computing in general that will be brought about by our work/ == Take note Mr Wolfram ... I think what we are doing will amaze even you (at least I hope so)
Just wrote a whitepaper on the Watson Ecosystem titled An Ecosystem of Innovation: Creating Cognitive applications powered by Watson. This should give you an initial overview of the Watson Ecosystem and building cognitive applications embedding the capabilities of Watson.
Feel free to leave a comment or post questions or discussions in the forum.
SridharS 110000N203 2,927 Views
A big day for us in the Watson Solutions as
"Today IBM announced that, for the first time, it will make its IBM Watson technology available as a development platform in the cloud, to enable a worldwide community of software application providers to build a new generation of apps infused with Watson's cognitive computing intelligence.
The move aims to spur innovation and fuel a new ecosystem of entrepreneurial software application providers – ranging from start-ups and emerging, venture capital backed businesses to established players. Together with IBM, these business partners share a vision for creating a new class of cognitive applications that transform how businesses and consumers make decisions.
To bring this shared vision to life, IBM will be launching the IBM Watson Developers Cloud, a cloud-hosted marketplace where application providers of all sizes and industries will be able to tap into resources for developing Watson-powered apps. This will include a developer toolkit, educational materials and access to Watson's application programming interface (API).
IBM partners that build Watson-powered apps in the cloud will be able to choose from two sources of data-driven content, to prepare their apps to uncover insights for users. App providers can use their own company’s data, or access the IBM Watson Content Store, featuring third-party content that offers data-rich resources that can fuel Watson’s ever expanding knowledge.
IBM will also connect its app-building partners with skilled professionals who can help advance their creative visions into market-ready products and services. IBM has committed more than 500 of its own subject matter experts as part of the overall program, in areas such as design, development and research. In addition, application providers participating in the ecosystem are expected to be able to connect with freelance professionals whose expertise can serve as a complement to their existing teams, from programmers to app developers."
Very proud to be an IBMer! For more information, see the press release.
SridharS 110000N203 1,455 Views
The University of Texas MD Anderson Cancer Center and IBM (NYSE: IBM) announced on Friday October 18, that MD Anderson is using the IBM Watson cognitive computing system for its mission to eradicate cancer. Following a year-long collaboration, IBM and MD Anderson showcased a prototype of MD Anderson’s Oncology Expert AdvisorTM powered by IBM Watson. The organizations will discuss their shared vision to leverage Watson’s cognitive computing power to help patients by enabling clinicians to uncover valuable insights from the cancer center’s rich patient and research databases.
This is a big milestone and is a proud accomplishment for us in the Watson team, and for all of us working in the cognitive computing space. You can find more information on A smarter planet blog (http://asmarterplanet.com/blog/2013/10/md-anderson-cancer-center-plans-to-use-ibm-watson-to-help-eradicate-cancer.html) or a video on youtube (http://www.youtube.com/watch?v=P7ShPS8JhRA)
SridharS 110000N203 781 Views
I want to welcome you to the Watson Users community. The Watson Solutions team would like to use this community to have a platform of interaction between people that are building and using Watson to build cognitive solutions. The intent is to use this blog to share technical musings on Cognitive computing, with a focus on Watson.
As a way of background, here's a video that provides a glimpse into how Watson works : https://www.ibm.com/developerworks/mydeveloperworks/blogs/video-portal/entry/ibm_watson_the_science_behind_an_answer?lang=en
I look forward to you joining the community and contributing to advance this exciting era of cognitive computing.