About monitoring online reputation software and decoding human languages

Par défaut

decoding_human_languagesTwo recent and really interesting articles deal with understanding online opinions :
Sentiment analysis: Understanding customers who don’t mean what they say, from Tech Republic
A computational model to measure how good (or bad) your puns really are, from Discover.

These articles highlight the importance of understanding online opinions.
In the past decade, online reputation has been one of the new concerns for companies. Indeed, it is a well-known fact that nowadays people use internet whether it be for shopping, reading the news, or using it for one’s profession.

As a result, the reputation that the companies have on line has become an important part of the communication strategy.
So they have to invest in order to monitor their reputation, otherwise they could face a bad buzz. To monitor e-reputations, specialized software seems to be the ideal solution, however to my mind it is not sufficient.

Even though software enables one to get online opinion, most software doesn’t analyze the writers’ profiles, explain what motivates them to express or determine if what they say is true.
In spite of technology research, human language remains hard to decrypt and the risk is to misread opinions.

However, there is software that exists which can be helpful to analyze opinion; either producing text analysis or producing quantitative indicators. Indeed, a large amount of e-reputation software makes online opinions easier to understand.

As all companies are concerned by online reputation, the software market is prosperous and will continue to be so. Thus, the professionals who decrypt online opinion play an important role, unless one day research succeeds in doing it.