The era of "big data" has come

Recently, countless experts and media have frequently referred to the "Big data" era has come, so what is the big data, only the literal understanding, the main feature of large data is "big": in addition to the traditional methods to obtain data, new sources of data are endless stream, to form a huge data flow changes over time. This data stream destroys the current data-monitoring and research industry and completely subverts people's understanding and imagination.

Why is big data so big? Where does the "big" come from? To answer this question, first of all need to understand the source of the data, the data mentioned here, for the sake of corporate marketing, we are in the narrow sense of the consumer behavior data. In the past, there are several ways to obtain such data: first, according to the product sales and distribution characteristics to understand; the second is to use the method of questionnaire survey to obtain consumer behavior and attitude of opinion, third, through observation or interviews to obtain consumer behavior data. In a word, in the previous methods of acquiring data, the researcher occupies the active position, the consumer is more passive.

However, in the age of large data, consumers have been active or unconsciously to their behavior information in a data way to stay somewhere, new media technology all the time quietly record our behavior information, the data formed only waiting for the conscientious to collect, analyze and judge. So to a certain extent, the most important data acquisition method in the large data age is to observe and record silently by means of new technology, and the total amount of information obtained in this way is much larger than the traditional method of collecting consumer data. In the process, consumers are more active (although they may not be aware of it), while researchers passively receive it.

The formation of a large data age, depending on the rapid development of new media, such as the daily production of micro-blog data content, far more than the data produced in the past. We can analyze a person's microblog content, to try to understand the basic demographic characteristics of this person, as well as its values, consumption, and so on, and even discover the various brand information the person is interested in on his personal microblog, and the attitude of the brand, the value of the big data is obvious.

The above content is only the continuous enrichment and expansion of the total horizontal data, in addition, the vertical expansion of data has also achieved a large data age. We can take the past survey or observation data as a cross-sectional study, each consumer corresponding to a series of data points, so that two axes together to build a two-dimensional plane map. But in the new media era, consumers all the time the behavior of information has been added a new timeline, the original two-dimensional plane into a three-dimensional space, this three-dimensional space of the total amount of data is certainly far greater than the original two-dimensional space, which is the new media led to the arrival of large data era proof.

Chinese ancient times to "sheep big" for the United States, today with the volume of data "big" is "good"? I'm afraid that's not the case. It's like looking in a haystack, and if you know where the needle is falling and the direction of the water flowing, then it's much more reliable to look at the needle in that range, and it's far more efficient than a little bit of searching across the ocean. The same is true of the big data age, what we really need is not the big data itself, but the intelligence information we want to get behind the big data, which is probably only one out of 10,000 of the total, all we need is one out of 10,000, and a lot of other information is rubbish to us. From this point of view, the larger the amount of data, the more spam information, these dazzling spam, the information we really want to deeply buried in the data stream.

This is today's public opinion monitoring industry facing the dilemma, I have seen countless public opinion monitoring service providers, proudly boast that they can monitor a large number of information sources, as well as every day or every minute can be grabbed a huge amount of information content, and some public opinion monitoring agencies began to compete with more extensive monitoring coverage. And this is precisely one of the misunderstanding of public opinion monitoring, in the past, the information sources of public opinion without understanding of the situation, without prior research and planning of the public opinion monitoring, the greater the amount of information obtained, the more spam.

So the public opinion monitoring is not taken for granted so simple, it needs to experience the monitoring source of the investigation and planning, monitoring source management and weight assignment, monitoring information and its attributes of the capture, public opinion data analysis and research, public opinion response strategy suggestions and response effectiveness evaluation such a series of process links, And not just the amount of information required to crawl is better. In such a process system, the early research and planning of public opinion monitoring source is especially important, which directly determines the rate of data acquisition and the validity and usability of grasping information. So different public opinion monitoring project, its monitoring scope should be different.

Another implication of the arrival of the big data age is that the cooperation between the public opinion monitoring business should be greater than competition. If the enterprise a in doing public opinion monitoring, Enterprise B is also doing public opinion monitoring, if two monitoring can according to the above said to do a good job in the early research planning, then the two acquired monitoring content should be relatively low overlap, so there is no fierce competition between the two. Conversely, if Enterprise A, Enterprise B, Enterprise C ... Many enterprises combine the information content database of monitoring acquisition, then form a comparatively complete industry database. Another way of thinking to understand, that is, if a number of enterprises can jointly establish a shared public opinion monitoring content database, each enterprise filtered out of their own content can be, so that both reduce costs, but also improve accuracy, the advantages of its multiple win can be revealed.

The big Data age obviously brings us a lot of convenience, but after a blind cheer and a moment of excitement, we need to seriously reflect on how to correctly dig out the greatest effect of large data, so that we can navigate the ocean of large data, rather than be lost by the torrent of large data.

(see version 11 for related topics)