New Delhi: Microsoft India has announced that Microsoft Translator will now offer real-time translation in five additional languages viz., Gujarati, Marathi, Kannada, Malayalam and Punjabi.

This takes the total number of Indian languages supported to 10, including Bengali, Hindi, Tamil, Telugu, and Urdu. With this, the service will now allow over 90% of Indians to access information and work in their native/preferred languages, making computing language-agnostic and more inclusive in the country.

Language translation is a core part of Microsoft products and services. Users can avail the benefits of AI and Deep Neural Networks-enhanced, real-time translation in all these languages while using Bing and the Microsoft Translator website, as well as the Microsoft Translator App for Android, IOS and Windows. The Microsoft Translator app can recognize and translate languages from text, speech and even photos. Microsoft is also rolling out support for these languages in the Microsoft Office 365 and the Swiftkey keyboard.

For organizations of all sizes, Microsoft provides APIs on Azure that they can use in their products to conduct their business across the country and globally in different languages. Businesses can easily integrate the Microsoft Translator text translation API into their applications websites, tools, or any solution requiring multi-language support, including e-content translation, e-commerce product catalogues, product documentation and internal communication among others. The Microsoft Translator service is part of the Azure Cognitive Services API collection of machine learning and AI algorithms.

Sundar Srinivasan, General Manager, Microsoft India (R&D) Pvt. Ltd., said, “We’re committed to empower every Indian and every business in India by bringing the power of AI into their daily life. Microsoft celebrates the diversity of languages in India and wants to make the vast internet even more accessible. We have supported Indian languages in computing for over two decades, and more recently have made significant strides on machine translation across languages. With this release, we are bringing in cutting edge machine learning tech to democratize access to information for everyone in India.”

Bringing Deep Neural Networks to language translation

Since early 2000s, Microsoft has been pioneering the traditional Statistical Machine Translation (SMT) paradigm to translate global as well as Indian languages. The incorporation of Deep Neural Networks into translating complex Indian languages brings more accuracy and fluency to translation. While SMT is limited to translating a word within the local context of a few surrounding words, Deep Neural Networks has the capability of encoding more granular concepts like gender (feminine, masculine, neutral), politeness level (slang, casual, written, formal), and type of word (verb, noun, adjective). Microsoft continuously iterates the translation models keeping in mind tech advancement and usage, and releases newer and improved versions to all its users in a transparent manner. You can learn more about Microsoft Translator here.

Microsoft and Local Language computing: Microsoft has been consistently working to provide local language computing in Indian languages for over two decades since the launch of Project Bhasha in 1998, allowing users to input localized text easily and quickly using the Indian Language Input tool. Microsoft also announced support for email addresses in multiple Indian languages across most of its email apps and services. Through its global Local Language Program (LLP), Microsoft provides people access to technology in their native language.

The post Microsoft adds five Indian languages to Microsoft Translator to help break communication barriers appeared first on Artificial Intelligence.

In the world of data science, one must excel in data creativity, scepticism and communication. Interesting insights from data analysis will have to be shared with senior-level decision makers and key stakeholders in a way that achieves maximum impact.

The ability to effectively present and communicate findings — by creating compelling stories — to an audience that could very well be outside the realm of scientific data analytics, is a determining characteristic of any successful data scientist. The good news is, finding and telling stories with data is an art, across disciplines, that anyone can learn.

Why is communication important?

The work of a data scientist extends beyond the database. As a data scientist, you won’t just spend hours poring over data sets, crunching numbers or interpreting key findings to reach conclusions. You will also have to:

• Explain how you arrived at a particular conclusion
• Justify rationally why you approached a problem in a specific manner
• Convey interesting insights in a way that gets people to think or act differently
• Persuade your audience that your results are conclusive and can be turned into something actionable
• Express why your findings are valuable and how they fit into the overall picture

That’s a whole lot of communicating.

A data scientist must be persuasive, compelling and credible, and the best way to do that is by ‘data storytelling’. According to Brent Dykes, author of Effective Data Storytelling: How to Drive Change with Data, Narrative, and Visuals, data storytelling is a structured approach for conveying data insights and incorporates three key elements: data, visuals, and narrative.

In this article, we will address the latter two elements. A narrative can be useful to explain to your audience what the data is indicating and why the insights are important. A visual — graphs, charts , diagrams— can be used to highlight insights in a way that spreadsheets and dashboards cannot.

A combination of narratives and visuals can be used to engage your audience and drive actionable insights that enable effective decision-making.

Narrative

Statistics can be overwhelming to understand, especially if you don’t exclude unnecessary information or add a narrative to the data. You want to present and convey your insights in a way that resonates with — and makes an impact on — your target audience. Here’s how to go about it:

1. Contextualise your data

If you were presented with a number ‘25’ without any layers of context, you wouldn’t have any idea of its true meaning. Is it someone’s age? Is it a percentage change in customer conversions? Is it the number of website visitors?

While displaying data in a vacuum doesn’t contribute to our understanding of its significance, contextualising it can place it within a larger setting or background to enable it to acquire its true and full meaning.

You must ask yourself what background information is relevant or essential to enable better understanding of your findings. This way, you can convert senseless data into real information — information that can be used as actionable insights. You must also:

a. Know your audience — Consider who you are presenting your findings and insights to. Is it a C-suite executive who is interested in overall business performance? Is it a marketing manager who wants to determine if specific social media campaigns are effective? Is it a sales team that needs to know if it can hit its quota this quarter? Your presentation and data visualisation need to be framed around the level of information your audience already has and what it wants to have.

b. Relate to your audience— The Harvard Business Review conducted a neurobiological study on storytelling and discovered that “character-driven stories with emotional content result in a better understanding of the key points a speaker wishes to make.” This means that decisions are often made on an emotional — and not logical — basis. By creating a concept-driven connection with your audience, you can increase audience engagement.

People hear statistics but they feel stories — Brent Dykes

2. Create a compelling narrative

The way you present information can shape your audience’s perceptions and behaviours. Powerful stories are framed within a bigger picture — they have a hook, momentum and captivating purpose. Thinking about how to frame stories can help you find compelling narratives that stick, educate people about complex issues and stimulate conversation on proposed solutions.

So, how do you frame your story and develop a powerful narrative?Consider the following questions:

• What is the purpose of your data analysis? Why you have interpreted this data in the first place?
• What is your message?
• What problem are you proposing to solve with your insights?
• What value does your findings add to your organisation?

The goal of data storytelling (with a narrative) is to stimulate critical thinking and discussion for business decisions — you should be able to trigger a response from your audience with a call to action.

3. Summarise your insights

Imagine you only had a limited amount of time or a single sentence to communicate information to your audience. What would you say? You only have to tell what they need to know. Thinking about your presentation in this way will enable you to be succinct, clear and concise with the story you want to tell.

Remember, good data stories include only as much information as is needed to directly address the objective of the analysis — any more than that and your audience will struggle to understand the point you’re trying to make.

Visuals

Data visualisations — graphs, charts, diagrams — are a great way to bring data to life. A powerful visualisation tells a story through the graphical depiction of statistical information. By creatively using data and statistics to show patterns and draw conclusions about hypotheses, you can capture and hold your audience’s attention.

Deliveroo, a popular UK-based food delivery app that has raised millions of dollars in investor funding, uses Big Data and machine learning to power food delivery. It’s VP of Engineering had this to say about data visualisation:

Graphs help our operations team understand and react to trends and agents all across the business are running queries on our data set 24 hours a day — Dan Webb

It’s also essential to understand when it most suitable to employ certain data visualisations. Crazy Egg uses a super cool info-graphic to demonstrate this:

Illustrating your conclusions with appropriate data visualisations is a dynamic form of persuasion. It allows your audience to discover patterns, see trends and digest large amounts of data at a faster rate.

In conclusion, the value of data increases when it fits into a story. Those who work with the data must be able to convey why it is fascinating or useful, else the value is lost. Your target audience must be able to relate to or react to your analytical solutions. By presenting data within a narrative (and aided by visuals), you can put a human perspective on factual data to augment an audience’s understanding of your findings and drive valuable insights.

The post Tell Stories with Data: Communication in Data Science appeared first on Artificial Intelligence.