Friday, 29 September 2017

Introduction to Artificial Intelligence - AI

Artificial Intelligence and Robotics

You need to have basic knowledge of AI. You don't have to be a Tech Pro like me to understand the principles of AI. So, in this article we would examine just the basics, no technical jargon. I would try to be very clear and simple.

We would look at the following:
1. What really is Artificial Intelligence
2. Application Areas of AI
3. Future of AI
4. Science Fictions versus Reality
5. Final Notes

1. What is Artificial Intelligence?
Artificial Intelligence is a field of study that researches how to make machines exhibit human intelligence. There may be other definitions but the key idea is  the focus on developing system that could think and make decisions based on their environments, mimicking the actions similar to humans.

2. Application Areas of AI
We would examine the 5 key  application areas of Artificial Intelligence although these may vary in some texts.

Natural Language Processing(NLP): This is an area of AI that studies how to enable machines understand and respond to human language. One aspect of NLP is Voice Recognition. The allows you to speak to an application, then the application understands and gives a response to your query. This may be to search for a term online or to perform certain actions. Examples of this is the Siri in Apple IPhone and IPad or the Cortana in Windows 10. We also have the voice search feature in Google.

Expert Systems: This is area of AI that researches on how to make machines exhibit the traits of a human expert. When a machine could perform a surgical procedure that is normally done by a physician, or when a system could track whether changes and make forecast, which is normally done by geographers etc. Such systems/machines are called expert system. For expert systems to work, it must have a knowledge-base which is normally fed into the systems memory.

Robotics: This researches on development of humanoids and other systems that could mimic human behavior and sometime have human appearance and features.

Neural Networks: This has to do with systems that could 'see' and recognize patterns. Used in area of bio-metrics. Examples are facial recognition systems, pattern matching. Others are handwriting recognition and fingerprint matching

Fuzzy Logic Systems: This are systems that performs regulatory functions and adjust parameters accordingly. An example is a system that monitors temperature in rooms in an apartment and adjusts the air conditioners. Another example is the autopilot in aircraft that monitors speed, altitude pressure and other parameters and keep the within acceptable threshold. 

3. Future of Artificial Intelligence
Would it be possible to build a system that could compete with or supersede humans in intelligence? The answer is no! The reason is that human nature and behavior is so complex that scientists have not fully understood all aspects of it especially when it comes to emotional attributes and relation with other humans. So how could scientists build something they don't even understand in the first place? That is the point. So how far could AI go? To answer this question, I would recommend you get the movie called 'Person of Interest' produced by Jonathan Nolan. After watching it, you could see how far AI can go. Another height AI could reach is the aspect of Internet of Things(IOT). This would allow machines to fix themselves and  to interact with each other with little human interaction.

4. Science Fiction Versus Realities
In movies, we see a number  of fictitious actions and stories, but the fact is that some of what we see is actually reality. 

Consider the following:
  • Cars can drive themselves (reality)
  • Humans can teleport to another location (fiction) 
  • A football team of robots to compete with humans (fiction)
  • A robot that can perform a surgery with little human intervention(reality)
I can't give an exhaustive list here, but I'll try to list as much as I can, say in another, article.

5. Final Notes.
Scientist in the area of Artificial Intelligence (including me) have a done much in and still have much to do. For every other person, we need to have the basic knowledge of how systems work, ranging from your computer to your phone, your automobile and so on. Here in my blog I would continue to explain technical concepts in simple terms so that everyone could understand.

I would like to thank you for reading and if you have any question, you can leave it in the comment box below.

Hidden Facts About Nigeria's Satellites - NigComSat

Since 2007, NigComSat have become a familiar word among Nigeria's technology You need to understand a bit of what this is all about, whether you are tech-savvy or not, whether you are a Nigerian or not.

Figure 1: Nigeria's Billion Dollar Satelites
I took time to gather this relevant information and I have presented it in a simple terms(no much technical jargon) to allow all my readers to understand. Especially for Nigerians, after reading, you will understand why some say the country still have a long way to go.

We are going to cover the following sub-topics in this article:
1. What is NigComSat
2. NigComSat-1 and NigComSat-1R
3. Any Benefits from these Satellites?
4. Final Thoughts and Questions Arising

1. What is NigComSat?
NigComSat is short for Nigerian Communication Satellite Limited. The company is described as "a company under the Federal Ministry of Communication". The company was incorporated in 2006 during the regime of Olusegun Obasanjo. The company had the mission to manage Nigeria's Communication Satelite and provide economic benefits of such to the country.
However, prior to the incorporation of this company, and initial contract to build a satellite has earlier been signed with a Chinese firm(between Nigeria's Ministry of Communication and a State-owned Chinese firm called CALT)

2. NigComSat-1 and NigComSat -1R
The Satellite being built by the Chinese firm was designated NigComSat-1. Remember that there has been prior three Satellites launched by Nigeria: NigeriaSat-1, NigeriaSat-2 and NigeriaSat-X (we would talk about this on a later article).
The satellite was finally completed in China, and launched from China in May 13, 2007. It was build to have a lifespan of 15 years.  This Satellite failed after about 17 months.

Reasons Why  it Failed
In simple terms, we can say that the power supply failed, so the Satellite could not receive power and so all the modules started failing until it could no longer be tracked. The technical explanation is that there was some problem with the solar arrays that supply power to the satellite. Recall that Satellites are powered by series of solar panels which in turn charge up power storage units within the Satellite. For NigComSat-1, these solar panels either malfunctioned or stopped working. The last contact with the Satellite was on November 11, 2008.

The Launch of NigComSat-1R
NigComSate-1R was lunched in December 19, 2011 as a replacement for the NigComSat-1. The contract for this second Satellite was given to the same Chinese firm that lunched the failed one. This contract was signed, about 5 months after the failure of the first one.

Any Benefits From These Satellites?
Let's take a look at the initial reasons why the satellite was launched. One of the reasons is to improve Nigeria Telecommunication and ICT Sector. Over the years, Nigerian firms, both government and private have always subscribe for data and other related services via external(non-state owned) satellites. This was expected to change with the launch of NigComSat. Additionally, cost of Internet services was expected to drop to the barest minimum. So have these benefits been acheived?. I think you know the answer.

Final Thoughts and Questions Arising?
 Since over six years of launch of Nigeria's NigComSat, we are yet to see any tangible benefits. Additionally, the cost of running these satellite remains immense as we are going to examine in another article.
So the question is, what is the track record of this Chinese firm that launched a failed satellite? How long have they been in the business of building satellites? Why were they given the contract again knowing that they have failed in the first one? Why do we continue to spend to maintain a satellite when there is no corresponding benefits to the citizenry? These are some of the question that begs for answers.

Monday, 25 September 2017

Big Data and Big Data Analytics Explained!

Big Data defined by the 3 Vs(Volume, Variety and Velocity 
The term "Big Data" have become quite common in the field of modern Relational Database Management and Data Analysis.
We would cover the following sub-topics:

1. Difference Between Data and Information
2. The Conventional Data Analysis
3. What is Big Data
4. Big Data Analysis
5. Final Notes

Difference Between Data and Information
What is Data in the first place? There is not perfect definition of data but we can say 'data is quantities gathered from measurement made in the environment where such quantities could further be analysed to produced useful information' - Kindson Munonye.

Examples of data could include the following:
1. Temperature measurements taken several times during the day
2. Tabulation of scores of students tabulated at the end of the semester exam
3. Tabulation of responses gotten from respondents through a questionnaire distributed
4. List of names and dates of births of all the students in a school

You can think of more example. All of this could be viewed as data. This means that they could be further analysed to produce useful information. Note this major difference between data and information: data is analysed or processed to produce information but information is not processed to produced data.

Let's now look at examples of information, the useful results or conclusion obtained when data has been processed.

1. The weather is hottest around midday (information derived from 1)
2. 80 percent of the students passed the exam. OR. Students did very well in the exam (derived from 2)
3. Majority of the students are teenagers

Unlike the data presented previously, these pieces of information would be very useful to any management in taking decision and making needed changes

Data Analysis (Conventional Data Analytics)
The Data Analytics we know is the used of statistical techniques to process data to yield useful informaiton. From the previous example I gave, you can see that conventional data is normally presented in tabular form of rows and columns.
To analyse this data you apply methods such as t-Test, ANOVA, ANCOVA, Correlation and Regression, MANOVA, Chi-Square etc. So you get to find mean, median, mode, standard deviation variance etc.
All of this works perfectly and are still very usedful today in the field of Research Statistics in Education.Some of the tools used for data analysis includes SPSS by IBM,  EViews and Statistica.

What is Big Data?
The question is, what if in gathering data from respondents, each respondents includes his passport photo? Or in taking temperature readings, you also take snapshot of the whether at the particular time? How do you analyze the images?
Definition: Big data is defined as very large and complex data sets that cannot be analysed using conventional data analysis methods.

So what exactly makes data qualify as big data? Lets consider 3 attributes. These attributes are knows as the 3 Vs, that is Volume, Variety and Velocity.

1. Volume: Big data sets assumes very large volume. The size ranges from hundreds of gigabytes terabytes and even petabytes.
2. Variety: Just and mentioned, big data contains data in different structures. Could be images, text, audio as well as data in video formats. Because of this, this kind of data is referred to as unstructured data as opposed to structured data arranged in rows and columns in tables.
3. Velocity: This refers to the high rate of growth or generation of the data. An example would be the growth of data in a social network site such as Facebook.

Example of big data
Example 1: A typical example would be users data from social network site. This would be made up of several billions of files of different format/
Example 2: Millions of emails stored in public mail server such as Yahoo together with to attachments to these emails.

Big Data Analytics
Since big data does not always have a well-defined structure, it would not be possible to used conventional data analysis tools to analyze them to yield needed  trends and information.Some of the big data analysis tools are:
MapReduce: A software framework for analysis of unstructured data
Hadoop: Developed by Apache is a framework used for processing of distributed data.
Hive: An open source analysis tool for querying and processing big data

Final Notes
The field of data analytics is evolving and so is data analysis technology improving continuously. So for folks engaged in Research Data Analytics, this is time to improve your skills in the area of big data analytics. Would there be a time when conventional data analytics would be completely replaced and irrelevant? I don't think so. But its necessary to move along with the trend and keep yourself up to date with the latest trend in Data Analytics.

Datarmatics Research Group

I'm going to tell you about this very important group you need to know about if you are member of the academia or engaged in any form of research or if you are a lover of Technology.
The name of the group as you know is the Datarmatics Research Group(DRG). It is also called Datarmatics Digital Research Support or similar designations.

Cross-section of Participants at the 6-day Workshop and Advanced Research Data Analytics
Cross-section of Participants at the 6-day Workshop and Advanced Research Data Analytics held from 4th to 9th of September 2017

About the Group
 The research group was created by Datarmatics Integrated Services Limited, a company registered in Nigeria as an ICT/Research corporation focused on creating enlightenment on the following areas:
1. Advanced Research Data Analytics
2. Computer Assisted Instruction (CAI)
3. Software Engineering, Application Development and Website Development#
4. Artificial Intellingence, Robotics and Advanced Programming
5. Database Administration
6. Video Production
7. Telecommunication/Computer Networking
8. Consultancy and Training Services#
9. Publications
10. Other ICT/Research related areas

The group is based in Port Harcourt Nigeria, with head office Location at:
No 2B, General Diriyai Street
GRA Phase 1
Port Harcourt,
Rivers State, Nigeria

In addition to this it maintains other consultancy units across the country.

Datarmatics Programs
Over the years, Datarmatics have organized workshops, webinars and seminars in a number of topics including Data Science, used of Software Tools in Research Data Analytics, Website Development, Basic Computer Programming etc.
One of these programs is the 6-Day International Advanced Research Data Analytics Workshop held from 4th to 9th of September, 2017 with participants from across the country and consultants from the Unites States, UK and Nigeria.

How to Join the Group
Many have asked questions on the requirements for joining the group. The easiest way is to participate in one of the workshops/seminars held every month. However as a prerequisite, you have have completed a program in a tertiary institution.
You can also request to join any of the Whatsapp group and get real-time updates from the team.

Final Notes
Datarmatics focuses on teamwork. This allows researchers to share knowledge and systematically build a knowledge base where every member of the team could leverage on for professional development