[Guest Post] Social Networks for Disaster Recovery

We are back from a short summer break with a special blog post; special because it is the first one that comes from a person outside the Agro-Know team. In fact, the author of this blog post is Mohammed Z. Al-Taie, a PhD candidate at the Universiti Teknologi Malaysia (UTM)Faculty of Computing (FC) – Soft Computing Research Group (SCRG). Even though the topic of the post is not related to agriculture nor any other green topic, the post provides an analysis on how the social media can be used in cases of disasters. Mr. Taie was kind enough to contact us and offered to publish some parts of his research through the AK blog; an offer that we could not refuse, as the analysis seems really interesting and thorough and on top of that, it focuses on the use of social networks (which we frequently use mostly for dissemination purposes) for alternative but useful purposes. You may find the post of the author below, without any revisions:


Almost every day, world communities are impacted by some kind of natural disasters, such as earthquakes, forest fires, airplane crashes, hurricanes, floods and diseases, which take the lives of hundreds or thousands of people and also lead to severe losses in the economy.

Dealing with the damage brought on by a natural disaster could be a daunting and difficult task unless very precise procedures are followed. One of the significant living aspects today, the communications, become a big challenge in the middle of a crisis as people would begin seeking for critical information related to “what, where, who and how”.

The Internet has brought magnificent changes to the speed with which people and information would gather around disasters sites, and there has been an increase in the use of information and communication technologies (ICT) during mass emergencies which can support the distribution of the information throughout social networks.

Social networks (or social media) applications and services (e.g. Facebook, Twitter and Friendster) are increasingly playing a vital role in disaster response and recovery by providing response information before, during and after the disaster taking place. For example, social network technologies have been used to gather hazard information for the 2010 Haiti Earthquake, 2009 USA Oklahoma Grassfires and 2008 Sichuan earthquake in China. Aviation accidents are no exception.

The first fatal aviation accident took place in USA in 1908 and resulted in the injury of the pilot and the death of one passenger, while the accident with the highest number of passenger fatalities happened in Spain in 1977 and lead to the death of 583 passengers.

However, the Malaysian Airlines Flight MH370, which disappeared on March, 8, 2014, remains one of the world puzzles that have not yet been solved.

In this post, we will see how social network concepts have been used to investigate the mesoscopic and macroscopic characteristics of one of the aviation disasters, which is Flight MH370, and analyze the statistical and topological characteristics in order to demystify some of the many ambiguous aspects of that flight. More details on this can be found in the original study, “Flight MH370 Community Structure” by the article’s author.


Flight MH370 social network

In July 2014, the Malaysia Airlines Flight MH-17 heading from Amsterdam to Kuala Lumpur was shot down in Ukraine, near the Russian border. Since then, speculations have emerged that its precursor, Flight MH370 that went missing on March 8, 2014, was also destroyed by a missile. A huge international effort has failed, until today, to find any trace of the plane, despite the use of modern equipment, such as the unmanned drones, to scour the Indian Ocean seabed.

Many theories thrived to explain the demise of Flight MH370. Among them is the idea that the plane landed at a US military base in a remote island or it was shot down during military operations. Some experts believe that the passengers and crew died from suffocation before crashing into the ocean. Others suggested that the plane was hijacked and flown to Afghanistan or may have been hidden from detection under another airplane.

The passengers were from different spots on earth: China (153), Malaysia (38), Indonesia (7), Australia (6), France (4), India (5), United States 3 (including two toddlers), New Zealand (2), Ukraine (2), Canada (2), Russia (1), Italy (1), Taiwan (1), Holland (1) and Austria (1). They formed different groups, such as a group of prominent China artists, a group of top management employees of a US company, the crew members, different-size families, and so on. One of the largest groups was a group of more than a hundred Buddhists (not necessarily acquainted with each other) returning from Kuala Lumpur to Beijing after being in a Buddhism conference with more than 30,000 people from around the world.

The Flight network (based on the profiles that we have) consisted of a large number of dyads and a small number of triads, quadrants and so on. We handled the network as it had no isolates because the aircraft network is considered like one-integral network since any individual onboard can interact with any other individual (except for special cases such as the pilots, children, disabled, etc.).


Fig. 1 MH370 Community Structure

Fig. 1 MH370 Community Structure


Through the use of community detection methods, we found that the Flight social network consisted of three larger communities: the Chinese artists, the delegation of the Freescale Semi-Conductor Company and the crew. Smaller communities included six and five-people families from China, a four-people family from Malaysia, a French family, an Australian family, a group of tourists coming back from a trip in Nepal, a group of Chinese individuals who used to work in Singapore and so on.

We utilized social network analysis, which is a research approach that analyzes the structures of social networks, and the relationships among its members, to analyze the mesoscopic and macroscopic features of the Flight community.

In this air travel community, there exist a large number of small size communities compared to a small number of large size communities. This is has been proved also for biological, information and communication networks to name a few.


Fig. 2 The chinese artists community Structure

Fig. 2 The chinese artists community Structure


Edges connect communities of smaller size, and also connect communities of smaller size to those of larger size. However, in the first case, they are more common, because the number of small-size communities is much greater than the number of big-size communities. This will help us to comprehend the degree at which links connect communities of different sizes. It also emerges that not only the smaller communities are well connected but also the larger communities.

Through the analysis, we found that the network density, which provides some important network features such as the speed at which information diffuses among passengers and the levels of social capital or social constraint of nodes, is relatively high.

We also found that the average degree, which is used to describe the structural cohesion of the network, is high which reflects the cohesive nature of the network.

Also, in searching for the network diameter, which indexes the extensiveness of the network, i.e. how far the two furthest nodes are from each other, we found a path between Catherine Lawton, 54, Australia and Yan Xiao, 27, China.

The network analysis also reveals eight nodes (passengers) that are not directly connected to the rest of nodes in the network, but only through their own sub-network members. These nodes are: (a) the two aircraft pilots, because usually, pilots can only be accessed through the rest of the crew members. (b) five toddlers who could interact only with their family members and (c) an old visually-impaired woman, who could only be interacted with through either her husband or one of their two friends.


Fig. 3. The Freescale Company Community Structure

Fig. 3. The Freescale Company Community Structure


The study also reveals a number of interesting findings such the validity of the “Weak Ties” concept, which was conceptualized in the 70s of the last century and meant to depict the importance of the weak ties that connect communities that otherwise become loose. These ties are necessary for information diffusion between socially separated components. As we can see in the original study, the three larger communities are fully connected, while they are partially connected to the other sub-communities (through weak ties).


Fig. 4 The Crew Community Structure

Fig. 4 The Crew Community Structure


Family members say they believe there has been a cover up, and they are ready to pay anyone who helps to reveal the real fate of the aircraft. Also, some of the passengers’ families have started receiving compensation payments. After all, there is much more to discover and we are waiting for more secrets to be disclosed.


About the author

MohammedMohammed Z. Al-Taie is a PhD candidate at the Universiti Teknologi Malaysia (UTM)Faculty of Computing (FC) – Soft Computing Research Group (SCRG)Previously, he was a teacher at Al-Salam University College in Baghdad, Iraq. He holds master’s degree in computer science and communication from the Arts, Science and Technology University (AUL) in Lebanon and bachelor’s degree in computer science from Al-Mustansiriya University in Iraq.

Before joining college as a teacher, Mr. Taie had more than 10 years of experience in IT-related projects. His fields of interest include machine learning, social networks, recommender systems and advanced web technologies. He published a number of studies in fields such as social networks, e-government and e-commerce, in addition to one book about social network analysis.


We would like to thank Mr. Taie for offering to publish his research outcomes through the AK blog. AK does not endorse the views and opinions expressed by the author in this post.


  1. The study is a good one. It can rightly be applied in the field of agriculture to study social interaction among farmers in different settlements.

    • Indeed, it would be interesting to see a similar case study or report applying this methodology in the agricultural context. Please let us know in case you are aware of anything related to that!


Leave a Comment.

Can you prove that you are not a bot? * Time limit is exhausted. Please reload CAPTCHA.