Theme 3. Research and Theory

"Computers & Education" is a technically-based journal about how computing can be used in education. The papers concern cognition and educational training with a scope that covers any kind of learning that is linked to technology. The journal has an impact factor of 2.775 and a 5-Year impact factor of 3.305.

 From humor recognition to irony detection: The figurative language of social media Journal: Data & Knowledge Engineering Impact factor 1.519
The paper is about how irony and humor can be detected and processed automatically. It clearly describes what humor and irony are and how it would be possible for a computer to recognize them. They come up with four different ways a computer can compare different text with each other. It is through the patterns: polarity, unexpectedness, emotional scenarios and ambiguity. They describe what they mean by this and create formulas that they later test.

The test is made with Twitter texts and hash tags. Then the computer compares the texts with different tags. E.g. the #humor-text is compared with #politics-, #irony-, #technology-texts. It compares 50 000 texts and after each comparison the computer learns. The computer uses different methods by combining the patterns: polarity, unexpectedness, emotional scenarios and ambiguity when it compares which leads to different results. You could criticize their data usage, because only because something is tagged as #Humor it does not mean that it is humor.

1. What the term theory really means is crystal clear to any scientist but for a normal person it can be difficult to understand. Very often the term theory is used incorrectly and confused with other terms. A scientific theory is highly regarded scientific knowledge and not a guess. A theory contains evidence, data, hypotheses etc. and they are all connected through logic and reasoning. A scientific theory can also be tested. If a theory is no longer valid because of new data, bad evidence, incorrect logic or something else the theory can be remade.
   
2. Design and action what best describes the theory of the paper "From humor recognition to irony detection: The figurative language of social media". The authors first takes up the issues of the problem and forms a solution. Then they create a computer system that compares Twitter texts. It consists of computation tools that learn how to determine what kind of text it is with help of their own created language models (patterns). I think it fits best with the design and action theory because it describes to the reader how to create this kind of system.
   
3. I think that the paper lacked explanation of the results. The theory is only telling you what you need and how to build the system to reach certain results. It is difficult to further develop the system since there is no description of why the results were like they were. It is any how compensated with a detailed description of what theories the system is based upon.