COM0014 Blog #2– Storytelling and Communication Styl

From this week’s lecture I learned the importance of format for storytelling. Similar to engaging into social media for business, although it seems to be a very common thing just keeping posts, there are many things to be care about when we actually doing it. In the following passages I will give my opinions to them.

It is vital to form a particular way for storytelling for grammar and tone. Just like the first glance at one’s Instagram account page, the first thing we notice would be the fantastic content and user’s format posting pictures of videos, some may post 3*3 to form a large picture in order to catch potential followers’ attention. One’s habit using grammar and tone for storytelling is similar to catch visitors’ eyes. 

Besides, the structure for a passage is also important. Recently I started reading the Economist, one thing I noticed immediately is that those articles’ structure are very clear. Authors tend to express their opinions for each paragraph in the first sentence and Usually they would lay out the whole point in the first paragraph which is very convenient for readers to improve reading efficiency.I think that is the thing we should learn from.

Engaging into comments with others is another important thing we should keep in mind. One evident benefit is that by reading comments we can get some advice from visitors which would improve yourself and followers’ experience, which is vital for building loyalty. Furthermore, the comments will let us know what they need. However, similar to writing stories, it is important to care for tones when replying to visitors in the way they prefer.

Blog #1 Amazing tools for social media monitoring and News Sources

Nowadays various methods for social media trend monitoring and sources for news and updates have popped up with the development of technology. Among them, there are some tools which I particularly prefer due to the convenience and their fast speed for delivering information to users. Today I am going to introduce them to you.

Google Alerts

Google alerts is one of the products produced by google in aim for delivering new interesting content to its audience. I found this amazing tool from the lectures (COM0012) I attended last semester. It has many selections on setting frequency sending news to users from at most once a day to at most once a week. Users would only need to input any keywords of what they are interested in including their own company, related industry news or even updates for competitors. Once everything is set up, users will receive emails for those new content. Below is an example of my settings on google alert:

You can even get a summary about any contents you or your email are mentioned on it. Besides, it is totally free. So it is very nice to the social media trend monitoring beginners like me as I do not have to cost too much over it.


This is another ideal tool but will cost a little money. On this site you can track for your own brand, a campaign or even any event of the industry and performing market researches. There is a selection on the dashboard page which enables you to decide to track from Twitter, Instagram, News, Forums and Blogs or not. It is very handy on getting a clear understanding about what is going on. I have to say Hootsuite worth every penny. 

Sources for news

Kaggleblog is one of the source I use to receive updated information for subjects that I am interested in. As it will post new winners’ experience and thoughts for each competition and introductions of new published programming packages, which is helpful for me to broaden my views. 

Another is news source is Linkedin. Right now I have decided to change the job for higher position in order to improve myself, I have to watch the industry’s news so that I could have a better answer for interviews.

So far these sites work perfectly fine for me. Hope these sites would be helpful to you.

#Socialmedialistenning #Socialmediamonitoring

Blog #1 Amazing trip

It is pretty awkward that my last vacation was about 6 years ago because after that I had to focus on my study and finding internships and jobs. On my last vacation, I went to Sichuan province in China as it has the best spicy cuisine there as well as cultural monuments. I got there in the midnight at summer, by the time I got there I smelt the smell of the hot pot. I went to several local famous hot pot restaurant for all the lunches and dinners. The taste exploded immediately after my first bite of the meat. To be honest, it wasn’t that spicy as I thought, but the numbing made me have cups of ice water to regain the feeling. 

Although it was unfortunate that the day I visited Giant Panda Breeding Research Base was raining, still seeing those lovely pandas in real life were great fun. 

What impressed me most was that on the fourth day I went to Mount Qingcheng, which is one of the birthplaces of Chinese Taoism and one of the famous Taoist mountains, surrounding by pines and cupressuses. It was almost dawn when I got to the Master Hall. There were a Taoist priest and a believer talking something while burning the joss sticks before the praying. Their every movements were full of peace which made me be covered with relax, confidence and hope. I took that picture from a distance as a memory. Then I started to wonder if that is the main function of religions, to make people peaceful and hopeful for their life. Although I come from China, I’m not a atheist because there are many things that could not be explained by science and it seems understandable if using religions to explain it. In addition, I think that religion also constraints human mortality to prevent us doing harmful things to others and obeying the morals. 

This vacation was worthwhile as not just it made me relaxed, but also enabled me to consider things which I seldom to think about. I really hope that I could have more chances to take vacations, but in short term it seems impossible not to mention such situation due to COVID-19. How was your last trip? I’d like to hear from you from the comments. 

#Trip #Panda

COM0011 Blog #4 Introduction to SEIR model

As the situation seems getting worse in Ontario, everybody is staying at home for safety. After searhing some documentations and references( I have some ideal to build a model predicting the amount of people get infected of COVID-19. The model I’m going to use is the SEIR model.

Starting with the introduction about the SEIR model, in this model we have four parameters. S is for susceptible, which means those not get the disease but with high possibility to get infected due to the lack of immunity after conntacting with infected people. E means those who have contacted with infected person but without the ability to infect others. I, of course, means the number of infected person with the ability to infect two type of person mentioned before. R means those who have immunity or are quarantined due to recovery.

For the coefficients, r is the number of infected person, β is the virus’s infection ability, γ generally is the reciprocal of the course of disease, α is the the probability of morbidity of the latent person.

So the ideal is that when time, t, starts to change, the amount of S also changes, when t equals to 1, that means the daily changes to S. Similar to E, I and R. In the beginning, S is close to the total popultion, then slowly infected to E. The amount of S decreases per day is determined by the number of people exposed to the disease (r), the coefficient of infection (β), the proportion of people affected (I / N), the proportion of susceptible people (S / N), the total number of people (N).Then the number of changes in E per day is the number of S converted into E per day minus the number of E converted into I per day. The rest of I and R are the same.

The difficulty of constructing the model here is because of r, β and S, E, I are hard to decide, as some of the patients who get infected without knowing the truth and treat it as a heavy cold, that’s where I can’t move forward right now, but there might be some chance to get the right model by using the current data we have with the adjustment of coefficients.

Some Interesting Projects from Kaggle


If you are feeling frustrated during the self quarantine, why not try to do something you never done before? Here I would like to recommed you some interesting projects from kaggle which are for beginner level and very easy to finish. It’s not like the jigsaw puzzle or lego which requires you to wait until it’s delivered and any money to spend on them. As long as you have a laptop you can start it right way.

  1. Hotel booking demand

This dataset gives you information for reservation, length of stay, number of people to check in and so on for you to dig about. Inaddition, this data set doesn’t contain any obvious industry backgrounds. It is good for you to perform analyzing on users’ routine behavior or comparing the reservation needs and occupancy rate between city hotels and holiday hotels, the best time to book a hotel in a year or even using the Logistic regression to make a prediction on hotel reservations. The data has 32 colums in total and 120k rows so it is not so big which would require the machine to run the script the whole day

2. US Accidents

This data set contains all the traffic accident over 49 states in the U.S. including the features such as the loaction of the accident, weather, temperature and so on. Similar as mentioned above, this data set also does not require obvious industry background so anyone can perform analyzing on it such as finding the factor driving to the accident result, making predictions on where the accident would happen. In addition, as this data set collected the data from 2016 to last December, so it is also highly recommend to do the data visulization to state the summary of the accidents over these 4 years.

The whole point of this artical is trying to find something for you to try if you are feeling frustrated instead of hanging outside. Anyway, stay home, stay safe.

COM0011 Blog #2 How to use LeetCode efficiently

Those many who are struggling on solving the problems on LeetCode might have the same quesition about how to solve the problems more efficiently. Here I would provide my advice based on my experience, background and understanding.

The first advice is that don’t satisfy with figuring out the problems, at least solve 20 to 30 problems for each category. For example, tree, list, binary search. The more you put time on those questions, the more experience you would gain when facing similar situations.

The next one is the common thing. Keeping patience and perservance. As there are about 20 categories and let’s say we have to do 20 questions per category, then that would be 400 in total and each question would take much time. Suppose you only have 3 hours to study everyday, then it would take about 1-2 months.

And the last one is that, for the first time doing the questions, if you can’t think any idea in 10 minutes, then take a look at the solutions posted by others. Looking at others’ solution is not a shame thing as no one was borned as a master on writing the codes. For the second time, try to write the code without looking at others’ solutions in 1 hour and pass the test.

All the things I mentioned above are to help you improve your own ability on writing the code and solutions and hope my advice would actually work.

COM0011 BLOG#1 Why peole stop posting personal actvities and thoughts on social media?


I used to post a lot of my personal activities and thoughts on social media years ago and suddenly stopped because of no reason. To my surprise, my friends did the similar thing. This is an insteresting phenomenon why people stopped posting personal activities and thoughts on social media.

There is a research paper from Computers in Human Behavior constructed and verified that there are several factors that affect this phenomenon, which are: privacy, confidence, self-efficacy and helpfulness. And the relationship among self-efficacy, the feeling of helpness and this phenomenon there is a postivie relationship. This means that the higher the self-efficacy and helpfulness the user feels, the more he/she would like to post something on social media. As for the confidence, however, there is a negative relationship between it and the phenomenon.

The write explained this conflict as the user who feels that self-efficacy and feel that social media help is higher, the use of social media may also be at a higher level. As a result, the user will start feeling more competition and it would be eaiser for them to burnout.

From my perspective, this can be explained as in the early stage of social media development, the frequency of users’ posting on social median increased with the gain in confidence in using it. After reaching a point of peak, users start care about their privacy and the management of their impression on social media, which create a burnout and thus people start reducing the frequency of social media use.

Reference: Bright, L. F., Kleiser, S. B., & Grau, S. L. (2015). Too much Facebook? An exploratory examination of social media fatigue

FACEBOOK: Why peole stop posting personal actvities and thoughts on social media?

TWITTER: Why peole stop posting personal actvities and thoughts on social media?

COM0012 Assignment #1


Hi all, in the beginning of this blog, I’d like to talk a little bit about my self and the reason I chose this topic. My name is Shirun Liu and I’m a data analyst in a consulting company and trying to improve myself in order to be competitive by doing the algorithm question on Leet Code. By doing those questions I have a little experience now and I want to share with you guys, I would be appreciate that if this can be an advise or the hint for you if you’re facing. The question I’m going to talk about today is Add Two Numbers II

Introduction for the question

In this question, we are given two non-empty linked list, the most significant digit comes first and each of their nodes contain a single digit. Add the two numbers and return it as a linked list.

For this qustion, probably the easiest solution would be changing the linked list into numbers, adding these two and then reversing it and changing it into liked list in the end.

However, it takes too much time when there are two large linked list, so I tried another way to approach


First of all, we have to be familiar what is linked list. According to Wikipedia. In computer science, a linked list is a linear collection of data elements, whose order is not given by their physical placement in memory.

The most obvious advantage of a linked list is that the way in which conventional arrays arrange related items is different from the order of these data items in memory or disk. A linked list allows inserting and removing nodes at any position on the list, but does not allow random access.

So I’m going to use stacks to decrease its processing time and increase its efficiency. In the first I will creat two null stacks and take the element in order from two different linked lists and add them, then put the results into another new stack and in the last, we put the new numbers into the empty linked list.

I will also show how it works by attaching the code for you to be easy to get understand.