Social Sensing And Analytics

The term analytics tells it has to do with some data in "data analysis". The "Social" part in "social analytics" makes this slightly more specific, in that it has to do with data, but a specific type of data, such as: social , social , who follows whom on Twitter, Facebook, etc. And who is spreading rumors in a social of people.

Simply put, it (social analytics) is the study of people, consumers that buy and post their reviews, so that other potential buyers can be warned or informed before making a purchasing decision.

It is also a study of how information in a social network of people and who is at the center of communication. Who is being or creating the point, a point at which information starts really fast. We call these individuals influencers.

And of course there is a whole array of software programs and libraries and packages that allow to , and data and to information.

Social analytics all of these things. Social analytics by nature is a highly interdisciplinary area.

On one of the spectrum is this field of computer science: data mining, machine learning, where we focus on data , , the data and data from different platforms and analyzing and visualizing it using software programs and doing this all in a large scale and in real time manner.

On the other end of the spectrum are other disciplines such as social sciences, communication studies, psychology, marketing and so on.

Here we try to answer questions such as: What motivates people to form or break a ? What are some communication that make a community of people together for a long time, and so on?

These other of study provide context around what we analyze in computer science and more importantly, why? Why we analyze what we are analyzing.

Of the examples of how social analytics has changed the we live. One prominent example that is often talked about when people ask me this particular question, just because it is everywhere, is websites like Yelp and Tripadvisor, or social review sharing .

Almost everyone travels today, and all of us are super busy, and during the limited few days we are spending , we want to make sure that everything is perfect, from hotels to restaurants, to activities...

Here we are talking about going to a place we have never been . And we do not know there. Some 20 years ago, you would probably have some family members or friends or friends of friends that have been there... and you hoped that they would share some good information.

Look at how we are doing the information today. Go to any hotel search website. Next to each hotel you will probably find not just a star , but that are written by visitors of the hotel. Actual people that have visited that particular hotel.

Was it clean? Were the staff members friendly? Was it loud? Was the swimming pool well ? Did you find any bugs? Do the staff members speak English? Was it worth the money?

So much information is already available online and as you know, what is amazing about this, is we are not talking about a few people in our friends' , we are talking about and of others that we have never met in our lives, but who are offering useful information.

So how can individuals make of this information? Bed bugs found in a hotel? Ants found and hotel staff did not do anything about it? Well you away from this hotel and go look for another hotel.

In a similar way, when choosing the next restaurant to your spouse on a date you can go find a lot of reviews on websites like Yelp.

It does not have to be hotels or restaurants, but it can also be about electronics, piano tuning services, movers, services, and so on. There is so much social conversation online today, we as individuals can do our own research for better and .

All for free!

for businesses. Suppose that you are trying to open up restaurant. You probably have a good idea of who are your competitors and where they are in Singapore or globally. But do you know how or well received they are? If you see a long outside your competitors' restaurants, that is probably a good of how popular they are.

But what about your competitor's restaurant? What aspects of the competitor's restaurant that are so to so people? What aspects to people so well? For this again, you can simply go to restaurant review sites like Yelp or "? go where" or Tripadvisor for more detailed information.

The reviews will reveal what of the restaurants that won or lost people's . And this piece of will be crucial to you as you try to open a restaurant.

As for government, social is important. We saw a lot of chaters online about the Singapore elections in September. There was just a lot of buzz online. In my social analytics class at the school of information systems, we have our students social conversations from Twitter and other such as Instagram about the election.

We can not only who says what about the general elections in different parties. In some cases, we can also track the exact location of the user that made a particular post. On the voting day itself, from the morning hours when the voting started till 1AM - 2AM after the vote was done, the chatter in this massive online space went on.

Using social analytics, governments can listen to what the citizens are saying, what are their topics of or interest, and what is the overall . Is it positive or negative?

We have seen different shifts in the technology world: the of the Internet, Personal Computers, Mobile , social networking sites, and so on. And we have seen plenty of examples of how the technology is changing the way we live and do things. One of the areas, and sort of positive examples of the technology revolutionizing the way we live is education.

My recent work in social analytics is a one year long research collaboration with the NIE, the National Institute of Education in Singapore.

Let me share with you a description of what this research is about: imagine a classroom of students. We are not talking about a classroom in a (physical) building. We are talking about a classroom, where teachers sign-on to a website from their workplaces and students also sign-on from their homes or they are at the moment. Students are grouped together into small groups of 4 to 5 people, and each group is given a that requires them to in an online chatroom.

At any given moment there is a dozen or more chats going on at the same time. Now, if this was a physical classroom, one teacher would have to go around listening to what each group is discussing one by one. Unless you can copy and paste the teacher and have multiple copies of the teacher, it is impossible for the teacher to all groups at the same time. So when 2 or more groups are having problems forward with their problem solving tasks and need their teachers to intervene, unfortunately some groups will have to wait or go on without having help.

But in a virtual classroom, we can build and a computer program that can assist student groups. These programs act like the teacher, except that they are not human. But they can be trained to behave like the teacher and assist students like the teacher. They can any problems students maybe having earlier on by doing social listening.

The programs will listen to social conversations as they happen. As students type and discuss problems in online chats. They can move on certain words or phrases that indicate problem. Such as when students are stuck. Or when they do not understand the problem. Or when they are not agreeing with and they start arguing in a very unproductive way. Or, they are going on into a completely wrong direction with their problem solving task.

The way our programs do all these things is that they are built to understand human language.

In our research we with a generic American or British style English corpus. We went on to enhancing it by adding local languages such as Singlish, slang words, acronyms and emoticons.

We use python language and nltk package as the basis and went on to write our own package for additional functions such as checking for Singlish, slang words and so on.

We have just finished one of our research. The value added here from an education perspective is that we have the process of early detection of possible problems in student's problem solving and doing this in a highly -able manner.

This allows educational institutions to certain tasks that teachers would normally do, to computer programs.

My other more recent research is with the Nielsen . I cannot share too much information. But in a : imagine that we can listen to many social conversations across many social platforms such as Facebook, Twitter, Instagram... and a whole wide of blogs and websites including news sites, and being able to such as the of a newly released movie or TV show or consumer buying behavior. Not just overall globally.

But imagine that we can do this sort of prediction in a more context like Singapore among male population, among young people between 25 and 35 years of age. Imagine that we can do this very accurately using computer programs. Using machine learning algorithms. Companies that have this sort of information earlier on will be able to better design their marketing campaigns by focusing on specific .

And not just that. Imagine that we can get to a point where we can identify social . Individuals in a social network that are highly followed by and trusted by other people. Movie production companies can the social influencers to spread good news or information about their movies or shows.

So without going into too much specifics, the general idea behind the second project with Nielsen company is that we can do social listening to better understand consumption behaviors. And we can predict future behavior by looking at past chaters captured in social media.

I can only provide a general answer as can vary from website to website and from country to country. I must say that my response to this particular question in this podcast, is not and it is not intended as legal advice.

First things first. If you something on the web, it makes that something public. It is public information, and it can in someone's google search. Anyone can view it.

If someone sees your content and prints it or it and emails to its friends (or what not), well you have just made your content public. So you really can do about it. Unless you have copyright in your content such as photos. You do not want someone to take your photos and post them their own. You can add copyright statement or add on your photos. Note that copyright is valid in all countries that by the Behring convention. Currently there are 168 countries that are included here. Including Singapore starting in 1998.

Now if you post something behind some type of protection, for instance using your user identification, if you put something on a social network like Facebook after you using your user identification. The owner of the website may impose conditions and restrictions on the use of the data that you post. This is a contract between you as a user, and the owner or provider of the social networking website. The social networking site owner may use the data you post in a certain way. Other users, people who follow you on Facebook (or what not, they) can use the data you post in a certain way. You must refer to the 's TOS (Terms of Use) for details. Their is a completely legal way to share content: on Facebook you can click on "share". So that when the content appears in your feed, people following you can see that this message originates from someone else.

In Twitter, sometimes we see people take another tweet instead of retweeting it, giving the original poster the due credit. They will just create a new tweet with the same content. It is like stealing someone's quote. Total plagiarism right? But, it seems you cannot really copyright your tweet because of the 140-character limit is too short, and other reasons.

Now there is a difference between a person visiting a webpage and reading the content, and a computer program called a visiting the webpage and the content. There is what is called a robot exclusion protocol which basically says that crawlers should not grab content from any website that uses robot exclusion protocol.

When it comes to social media or social networking websites, some of the websites provide a public API. API stands for . Simply put, it tells you how to what data from a particular social media site. APIs will your data access ensuring that you, as a program, are able to access only certain types of data. Again, you would have to refer to the Terms of Use for what ever API to use for particular regulations and restrictions.

So basically there are 2 things. As a user posting some content onto a website, you need to understand who can access the content and how your content can be used by the website provider and by other users on the website. This can be found in the Terms of Use. Secondly, if you are a user of the website wanting to access other users' data, either by web crawling or through API, you need to refer to the Terms of Use again.

If you are crawling data from other websites for research , you need to make sure that you have gone through he IRB training as required by your respective institution. And you can only information. You cannot publish personally identifiable information. You would have to work closely with your institution, your schools, IRB office on this, if your research is dealing with a human topic. All this data we are crawling from the web: social conversations, social networks and so on... are about people, human subjects.

So please take a note of this as you are embarking on a research using social media data.