quiz 4 bis

3 programming languages for website - the front end: builder artist wizard

'The Builder' - HTML: Hypertext Markup Language Arrange the content: text, images, button.. . 'The Artist' - CSS: Cascading Style Sheets Add 'styles' to the content 'The Wizard' - JavaScript: dynamic effects Auto complete, popups... Browsers translate these three into what we see as webpages Languages evolve over time: HTML5, CSS3...

Related concepts

Gig economy, sharing economy "Pure" crowdsourcing business Open-source Online community

Usability

"the ease with which a user can learn to operate, prepare inputs for, and interpret outputs of a system or component" - IEEE Standard Computer Dictionary A key component in Human-Computer Interaction (HCI) A primary focus in the design of websites A "usability subculture" has evolved as a result of web technology Usability variations Usefulness, and user experience Usability engineering and usability method User-centered design

Conversion rates

A conversion rate is loosely defined as the percentage of website visitors who complete a desired action: Clicks beyond home page Membership registrations, newsletter subscriptions Sample downloads Sales inquiries, purchases etc. A high conversion rate depends on several factors The interest level of the visitor (marketing) Matching the right visitor, the right place, and the right time The attractiveness of the offer (marketing) The ease of the process: User interface Convenience Performance

Community Principle #1: Public Goods and Free Riders

A public goods are subject to: Nondimishability One person's use does not reduce the availability of the good for another Nonexcludability Individuals are not prevented from using the good because they have not paid for it Public goods are subject to free riding (individuals who use a good without contributing to covering its cost)

Crowdsourcing

A sourcing model in which individuals or organizations obtain goods or services including ideas, voting, micro-tasks, and finances from a large, relatively open, and often rapidly evolving group of participants When businesses provide enabling technologies that allow people - instead of paid employees - to create, modify, and oversee the development of a product or service Typically depend on computer mediated communication

What Is Artificial Intelligence?

Artificial intelligence (AI): a form of intelligent technique Grand vision Computer hardware and software systems that are as "smart" as humans So far, this vision has eluded computer programmers and scientists Narrower, more realistic vision Systems that take data inputs, process them, and produce outputs (like all software programs) and that can perform many complex tasks that would be difficult or impossible for humans to perform.

Basic tasks and business applications

Basic tasks and business applications Which type of ML? Sentiment analysis Social media monitoring (demo) Language interpretation & generation Translation (video), chat bots Speech recognition Voice assistant (i.e., Siri, Alexa) Many more!

Break down 'usability' Dimensions

Break down 'usability' Ecommerce website is where the customers can search and retrieve any information and perform other tasks like adding to cart, save for later, checkout, payment, tracking records and many more Dimensions Appearance & tech: simplicity, attractiveness, speed Content & structure: searching & navigation, product categorization, product information By page function Homepage Landing page Category page Product page Checkout

Expert Systems

Capture tacit knowledge in very specific and limited domain of human expertise Capture knowledge as set of rules Typically perform limited tasks Diagnosing malfunctioning machine Determining whether to grant credit for loan Used for discrete, highly structured decision making How it work? Knowledge base: set of hundreds or thousands of rules Inference engine: strategy used to search knowledge base Forward chaining Backward chaining

How to build a website? CMS

Content management system (CMS) A software that allows you to manage website content in a simple and interactive way Examples WordPress.org: the most popular solution; 43% of the web uses this solution Open source (free to use) HubSpot CMS, Wix Tiered pricing plans (many others) ... Now, how to build a website? Each website is essentially an application running on a powerful computer Get hosting service (server, network); get a domain name Manage the content

Types of crowdsourcing to solve business problem 5

Crowd contests Macrotasks Microtasks Crowdfunding Self-organized (as the others tend to rely on platforms)

Machine Learning

Different paradigm (comparing to expert systems) Expert systems are rule-based Machine learning (ML) is based on statistic models ML basics How we present the process of business decision-making with data? What's the decision? E.g., yes or no? ranking? other quantity? Also called 'output' Variables to describe other aspects (or features) that are relating to the decision x_1, x_2,..., x_n; also called 'input'

Ten Most Violated Homepage Design Guidelines

Emphasize what your site offers that's of value to users and how your services differ from those of key competitors Use a liquid layout that lets users adjust the homepage size Responsive design: webpages that can adjust to mobile devices Use color to distinguish visited and unvisited links Use graphics to show real content, not just to decorate your homepage In e-commerce, it means the photos and videos should be related to the products Use good quality photos and videos; optimize for loading speed Include a tag line that explicitly summarizes what the site or company does

Application - computer vision (CV)

Emulate human visual system to view and extract information from real-world images How can computer understand images? Transforming them into numbers! All the pictures are in pixels Basic tasks: Object detection: classification & localization Identify objects in an image, assigning labels and "localizing" the objects Segmentation: Understands every pixel of an image Mostly rely on DL

Types of AI 2

Expert systems Modern AI are based on machine learning Types of ML Deep learning and neural networks

Business applications of computer vision

Facebook's DeepFace can identify friends in photos across their system and the entire web Autonomous vehicles can recognize signs, road markers, people, animals, and other vehicles with good reliability Handwriting recognition: a drawing game with Google Cashierless stores: e.g., Amazon go grocery (video) Image (video) transformation: photo editing apps Medical image analysis

Firms use crowdsourcing as solutions...

Firms use crowdsourcing as solutions to business processes Primary value process (from value chain's perspective) Make the product or service Marketing & sell: social media marketing Service after the sell Support value process

Crowdfunding

Fund a project or venture by raising small amounts of money from a large number of people Provide ways of funding for: Start-ups and entrepreneurs Individuals with financial difficulties - P2P lending Creative work Types Reward-based Equity-based Donation Debt-based - mainly for P2P lending Crypto-based

gig & sharing economy

Gig economy/platforms "a labor market characterized by the prevalence of short-term contracts or freelance work as opposed to permanent jobs" Overlap with crowdsourcing Sharing economy

Ten Most Violated Homepage Design Guidelines continued

Make it easy to access anything recently featured on your homepage Include a short site description in the window title Don't use a heading to label the search area; instead, use a "Search" button With stock quotes, give the percentage of change, not just the points gained lost Don't include an active link to the homepage on the homepage

What's a neural network (NN)?

Mimic neuros of human beings

AI, ML, Deep Learning & Neural Networks: What's the relationship among these concepts?

Modern AI systems are built on ML models DL & NN are ML techniques Can be used for any ML problems: classification, clustering etc.

Managing online community

Modern business involves managing online community due to various reasons Customer relationship, branding 'Pure' crowdsourcing business Open-source software community Social networks Collective knowledge platforms

Deep learning (DL), or deep neural networks

More complex, with many layers of transformations of input data to produce target output Some believe these come closest to "grand vision" of AI

Crowdsourcing - crowd contests features

Only pay for the top solution Utilize internal & external talents Most effective when the problem is complex or novel or has no established best-practice approaches Creative work: e.g., design a logo

Open-Source Software

Open-source communities act as... Software vendors Development Maintenance Support User groups Standard setting & governance body How can OSS make money? Support & additive service Hosting: provide cloud version of an OSS; such as Red Hat Open-core: most popular way Like freemium; such as Canvas How is OSS different from crowdsourcing? Not the same; some OSS businesses are not pure crowdsourcing

E-commerce website - best practices

Overall: embrace the 'flat and modern' design; consistency over the entire site Homepage/landing page: Clear call to action (CTA): stand-out color but not intrusive Value proposition Category, product & checkout: visual hierarchy and good navigation Easy-to-use filters Use breadcrumbs Checkout: progress indicator & security payment badge, no distractions Accessibility When in doubt, use A/B Testing Most best practices are incorporated in ecommerce solutions Squarespace, Shopify, Wix, BigCommerce

Managing online community

Practical implications from three principles Evaluate your online community What's the public goods? Who are the active users or contributors? Design your online community features to promote desired actions from users Such as vote up and down, reputation system etc.

Applications - natural language processing (NLP)

Process voice or text command using natural human language Deep learning has improved the performance a lot How can computer understand natural language? Written: encoding sentences into numbers (from unstructured to structured) One basic method: bag of words Advanced method: word2vec Speech Break down audio into individual sounds, then converts sounds into digital format (such as frequencies)

'Pure' crowdsourcing business - Wikipedia

Product = encyclopedia 30 million articles in 287 languages Who competes with Wikipedia? How is Wikipedia different (better/worse) from Encyclopedia Britannica? Does anyone make money from Wikipedia? Why do people contribute for free

Crowdsourcing - pros & cons

Pros Solve problems more quickly Save time and money Build up customer data, get feedback Get more innovative and diverse ideas Accelerate certain processes Cons Less control over the process Risk of inconsistency Internal conflict Issues with confidentiality being compromised Poor quality entries

Applications - Recommender systems

Recommender systems Or recommendation systems are algorithms aimed at suggesting relevant items to users; also a subclass of information filtering system It's almost everywhere in the digital world! Social networks, app store, Amazon etc. How it works? Content filtering Based on similarity among content (i.e., products, movies and songs etc.) Collaborative filtering Based on past interactions between users and items Could use 'clustering' to group users

Limitations of neural networks and machine learning

Require very large data sets to identify patterns It's a black box! How system arrived at a particular solution often cannot be explained No sense of ethics, so may recommend actions that are illegal or immoral

Supervised learning Unsupervised learning Reinforcement learning

Supervised learning Trained by labelled data, that is data with both input and output How it works Example: regression (for continuous output) and classification (for categorical output) algorithm to forecast and predict Unsupervised learning A more 'hands-off' approach: only the input, but no output How it works Example: clustering algorithm to identify customers who are similar; dimension reduction when there are too many input variables Reinforcement learning The program learns from trial and error How it works Example: AlphaGo

Conversion rates & usability

Target inappropriate audience (marketing) Unclear marketing message (marketing) Slow page load (usability) Un-engaging look and feel (usability) Clumsy site navigation (usability) Ineffective presentation (usability) Inadequate selection (usability) No access to real time help (usability) Uncompetitive price Ineffective tools to assist selection (usability)

Microtasks

Tasks that can be divided into small units Normally skill independent; low-value tasks

Macrotasks

Tasks that can be done independently Require special skill Worker paid by task Ideal for SME Similar to outsourcing, but to a larger audience

Business decision - what type of AI to use?

The choice of AI technology should be driven by the business need What's your business question? Some may need more than one type of AI (or ML models) There are many ready-to-use AI models, especially cloud solutions Such as, Google Cloud Platform's machine learning APIs

two perspectives crowdsourcing

Two perspectives How can firm use crowdsourcing? Design and manage crowdsourcing platforms

Community Principle #3: Diversity

What is costly to one is cheap for others; what is low benefit or uninteresting to one is highly beneficial and desirable for another Questions & answers Software Development effort Community maintenance Multiple motivations support long term viability

see workshop 4

examples in ppt

look at

powerpoints/ workshops