Upload
charles-quincy
View
62
Download
2
Tags:
Embed Size (px)
Citation preview
IBM Watson Developer Cloud Services Overview
January 2015
IBM CONFIDENTIAL
IMPORTANT DISCLAIMER
2
IBM’s statements regarding its plans, directions, and intent are subject to change or withdrawal without notice at IBM’s sole discretion. Information regarding potential future products is intended to outline our general product direction and it should not be relied on in making a purchasing decision. The information mentioned regarding potential future products is not a commitment, promise, or legal obligation to deliver any material, code or functionality. Information about potential future products may not be incorporated into any contract. The development, release, and timing of any future features or functionality described for our products remains at our sole discretion.
The information contained herein is IBM Confidential and may not be used or disclosed without express written permission from IBM Corporation.
IBM CONFIDENTIAL
Personality Insights aka User Modeling (available in Beta now)
3
What is it? Personality profiling to help engage users on their own terms. How does it work? The Personality Insights service uses linguis9c analy9cs to extract a set of personality and social traits from the way a person communicates. The service can analyze any communica9on the user makes available such as their text messages, tweets, posts, email, and more. Users of the service can understand, connect, and communicate with people on a more personally tailored level by analyzing personality and social traits Example Use Cases • Analyze a customer’s twiEer stream to help a travel
agency decide between leading with a budget or luxury trip offer
• Improve customer engagement by personalizing and differen9a9ng products and offerings using customer’s public communica9on (e.g., twiEer feed)
IBM CONFIDENTIAL
Question and Answer (available now)
4
Understands question
Produces possible answers and evidence
Analyzes evidence
Computes confidence
Delivers response, evidence and confidence
Asks a question
Considers response and evidence
What is it?Direct responses to users inquiries fueled by primary document sources
How does it work?Interprets and answers user ques9ons directly based on primary data sources (brochures, web pages, manuals, records, etc.) that have been selected and gathered into a body of data or ‘corpus’. The service returns candidate responses with associated confidence levels and links to suppor9ng evidence. The current data corpora on Bluemix focuses on the Travel and Healthcare industries.
Example Use CasesHealthcare:
• What is a stroke? What is the cause of Wilson Disease?
Travel: • Where is the best place to stay in Prague?
IBM CONFIDENTIAL
Relationship Extraction (available in Beta now)
5
What is it? From unstructured text, Relationship Extraction can extract entities (such as people, locations, organizations, events), and the relationships between these entities (such as person employed-by organization, person resides-in location) How does it work? Based on statistical modeling, Relationship Extraction performs linguistic analysis of the input text, finds spans of text that refers to entities, clusters them together to form entities, and extracts relationships between them. Relationship Extraction models are domain-specific and work best with in-domain input. Example Use Cases The output of Relationship Extraction, namely entities and their relationships, can be used to build applications such as a semantic search engine, business and competitive intelligence tools, knowledge graphs, social media monitoring tools, etc.
IBM CONFIDENTIAL
Message Resonance (available in Beta now)
6 Visionary mobile protoype
What is it? Communicate with people with a style and words that suits them How does it work? The message resonance service analyzes draR content and scores how well it is likely to be received by a specific target audience. This analysis is based on content that’s been wriEen by the target audience itself such as fans of a specific sports team or new parents. Today, analysis can be done against people ac9ve in cloud compu9ng or discussions but future versions will let users provide their own community data. Example Use Cases Among people ac9ve in cloud compu9ng discussions, op9on A content is likely to resonate very well, op9on B poorly, and op9on C moderately well.
IBM CONFIDENTIAL
Concept Expansion (available in Beta now)
7
What is it? Maps euphemisms or colloquial terms to more commonly understood phrases How does it work? The Concept Expansion service analyses text and interprets its meaning based on usage in other similar contexts. For example, it could interpret “The Big Apple” as meaning “New York City”. It can be used to create a dic9onary of related words and concepts so that euphemisms, colloquialisms, or otherwise unclear phrases can be beEer understood and analyzed. Example Use Cases “drugs” can be expanded to:
start seed terms à motrin, aspirin, Keflex
post expansion à allegra, lisinopril, me\ormin, aspirin, equagesic, cime9dine, fiorinal, vancomycin, avelox, protonix, glimepiride, protonix, verapamil, norco, inderal, hctz, advair
IBM CONFIDENTIAL
What is it? Globalize on the fly. Translate text from one language to another. Customize the transla9on algorithm for your specific use case, or use a generic (news) one. How does it work? The Machine Transla9on service converts text input in one language into a des9na9on language for the end user. Today, bi-‐direc9onal generic transla9on is available between English and [French, Spanish, Portuguese]; uni-‐direc9onal generic transla9on is available from Arabic to English. More languages and customiza9on capabili9es coming soon!
Example Use Cases • A news agency is able to curate ar9cles from across the globe, and present them to a local audience in their na9ve language.
• An e-‐commerce website can enable a truly global marketplace.
• Travel apps can deliver on the promise of a fully guided, interac9ve experience.
Machine Translation (available in Beta now)
8
IBM CONFIDENTIAL
Language Identification (available in Beta now)
9
What is it? Iden9fies the language in which text is wriEen. How does it work? The Language Iden9fica9on service detects the language in which text is wriEen. This helps inform next steps such as transla9on, voice to text, or direct analysis. The service can be used in tandem with the Machine Transla9on service. Today, the service can iden9fy 15 languages – Arabic; Chinese (Simplified); Chinese (Tradi9onal); Cyrillic; Danish; Dutch; English; Farsi; Finnish; French; German; Greek; Hebrew; Hindi; Icelandic; Italian; Japanese; Korean; Norwegian (Bokmal); Norwegian (Nynorsk); Portuguese; Spanish; Swedish; Turkish; Urdu
Example Use Cases A building block for Machine Transla9on and future tech
IBM CONFIDENTIAL
Visualization Rendering (available now)
10
Visionary mobile prototype
What is it? Graphical representa9ons of data analysis for easier understanding How does it work? The service takes input data and graphically renders it as an interac9ve visualiza9on which can range from a common business chart to more advanced layouts. The visualiza9ons can be easily modified to match user needs, visual styling, and types of data being analyzed. Example Use Cases The service could represent neighborhood demographic data as mini pie charts showing income levels centered on geographic loca9ons on maps, or as tree maps that can switch from looking at income by age to house size or by educa9on level.
IBM CONFIDENTIAL
Speech-to-Text (available in beta early 1Q15)
11
What is it? Automa9c recogni9on of the words that are being spoken in a sound file, and presents the highest-‐scoring textual representa9on of that message. How does it work? In addi9on to conver9ng a raw audio signal into a best-‐guess of the words that are being spoken, intelligence about the relevant grammar / how language is used within a specific context is incorporated to generate a more accurate transcrip9on. Ini9ally coming out with a general (news) model for English; will eventually be able to tweak the algorithm for a specific use case (e.g., “Men in 9ghts” in the general domain, “Meningi9s” when customized to the Medical domain).
Example Use Cases • Transcrip9on of mee9ngs and conference calls. • Voice-‐control of applica9ons / embedded devices. • Dicta9on of emails. • Cri9cal building block for “Speech-‐to-‐Speech” transla9on.
IBM CONFIDENTIAL
Text-to-Speech (available in beta early 1Q15)
12
What is it? Generates an audio file that has a verbal representa9on of the input text – complete with appropriate cadence and intona9on, and the ability to customize the pronuncia9on of specific words. How does it work? Generates an audio file that has a verbal representa9on of the input text – complete with appropriate cadence and intona9on. Ini9ally coming out with a general (news) model for English and Spanish; will eventually be able to customize the pronuncia9on of specific words, and tailor enuncia9on appropriate for a given domain. Example Use Cases • Read texts / emails aloud. • Assistance tools for the vision-‐impaired. • Transmit informa9on from a computer to a human without using a screen.
• Cri9cal building block for enable “Speech-‐to-‐Speech Transla9on”
IBM CONFIDENTIAL
Concept Insights (available in beta early 1Q15) What is it? Locate relevant documents that may not directly men9on your query. How does it work? The Concept Insights service links documents you provide with a pre-‐exis9ng graph of concepts based on Wikipedia (e.g. 'The New York Times', 'Machine learning', etc.). Two types of links are iden9fied: explicit links when a document directly men9ons a concept, and implicit links which connect your documents to relevant concepts that are not directly men9oned in them. Users of this service can also search for documents that are relevant to a concept or collec9on of concepts by exploring the explicit and implicit links. Example Use Cases • A legal firm could u9lize this service to iden9fy cases
which may be related to the case in ques9on • Improve engagement on any external website
IBM CONFIDENTIAL
Visual Recognition (available in beta early 1Q15)
14
What is it? Analyzes the visual content of images and videos to understand their content without requiring a textual descrip9on. How does it work? The Visual Recogni9on service enables you to analyze the visual appearance of images or video frames to understand what is happening in a scene. Using machine learning technology, seman9c classifiers recognize many visual en99es, such as semngs, objects, and events. The service applies these pre-‐learned models to imagery that you have uploaded to the service and returns a score for each image for each model, indica9ng the likelihood of that visual element being present in the image. Example Use Cases • Audio-‐visual indexing and search of media archive • Automa9c “Smart Album” genera9on • Marke9ng data analysis from social media -‐ For e.g., mine Pinterest images and perform segmenta9on
IBM CONFIDENTIAL
Tradeoff Analytics (available in beta early 1Q15) What is it? Helps make beEer choices under mul9ple conflic9ng goals. Combines smart visualiza9ons and analy9cal recommenda9ons for tradeoff explora9on. How does it work? Tradeoff Analy9cs helps people make beEer choices when mul9ple conflic9ng goals come into play. Tradeoff Analy9cs uses Pareto filtering techniques in order to surface out only the op9mal alterna9ves across mul9ple criteria. It then uses various analy9cal and visual approaches to help the decision maker analyze the tradeoffs within the op9mal set of alterna9ves helping them select the op9on that best meets their preferences and priori9es. Use Cases • Front end sales portal using Tradeoff Analy9cs to empower customers' decision capability and collects data on what customers deem most important to product purchases
• Wealth management using the technology to help choose mutual funds