How customer service will be changed in the era of big data_Data Analyst Exam
Recently, "big data" has replaced "cloud technology" as the hot topic of the new technology, all kinds of "big data" books are endless, articles are numerous, as if you don't have something to do with "big data", as if you don't have something to do with "big data". "Big Data" books are popping up all over the place, and the articles are full of them, as if you don't have something to do with "Big Data", you're out of the game! I have also dabbled a little in these articles, but I feel that there are many people who hang on to the false name of "Big Data", but few people who really know it. In order to make it easier to understand the connotation of big data? Allow me to start with a brief introduction to the definition and background of Big Data.
McKinsey's report defines it this way: Big data refers to a collection of data that cannot be captured, managed, and processed within a certain period of time using traditional database software tools. (Big data refers to datasets whose size is beyond the ability oftypical database software tools to capture, store, manage, and analyze.)
How did the concept of big data come about? In May 2011, EMC hosted a conference on the theme "Cloud Computing Meets Big Data," where the concept of "Big Data" was first introduced; in June, IDC's annual Digital Universe study, "From Chaos to the Universe," was sponsored by EMC and produced by IDC. In June, sponsored by EMC, IDC compiled the annual digital universe research report "Extracting Value from Chaos" (Extracting Value from Chaos) was released; immediately after, IBM, McKinsey and many other foreign organizations released "Big Data" related research reports to be actively followed up.
From the background, we can see that EMC (the world's largest provider of external storage hard drives) is the mastermind behind the concept of "big data," and by doing so, of course, he wants to sell more hard drives. This kind of soft advertising hype not only did not cause resentment and spit, but recognized and accepted by all sectors of society, but also with its social background is inextricably linked. As the cost of data generation has fallen sharply in recent years, the amount of data generated by humans is growing exponentially, of which more than 80% is unstructured data that cannot be processed by traditional databases. How big is this data? According to IDC's monitoring, the world officially entered the ZB era in 2010, and it is expected that by 2020, the world will have a total of 35ZB of data, and if all the 35ZB of data are burned onto a CD-ROM with a capacity of 9GB, the height of its stack is equivalent to three round trips between the Earth and the Moon. ...... In such an intuitive analogy, other languages can be used as an example. In the face of such an intuitive analogy, other words will pale in comparison!
Maybe you'll say that the phenomenon of big data goes without saying, we've seen it before, it's not just that the data is big, what substantial impact can it bring to our society ah, or how come I don't see its application? On the application of big data, I will not repeat here, the market a variety of "big data" book has talked about a lot of cases. I just want to say that "analyzing history can provide insight into the future". A few years ago, those who said that "cloud technology" is still far away are uploading their documents, photos and videos to "iCloud", using "iCloud", "iCloud", "iCloud", "iCloud", "iCloud", "iCloud", "iCloud" and "iCloud". They are uploading their documents, photos and videos to "iCloud", using "Sogou Cloud Input Method", and logging into websites such as Dropbox, Yelp, Zynga, etc. (which are being hosted on Amazon's "cloud platform"). ...... The "cloud" floating in the sky is no longer "inaccessible".
Big data will bring significant value to healthcare, public **** management, location-based services, retail, and manufacturing, according to the McKinsey Global Institute's study, "Big data: The next frontier for innovation, competition, and productivity. For example, for the U.S. healthcare services industry to create value of $300 billion per year, about 0.7% annual growth rate, for the U.S. manufacturing industry up to 50% reduction in product development and assembly costs. This McKinsey report spells out in detail the benefits of big data for a wide range of industries, and is recommended reading; I'll just talk about the impact of big data on the customer service sector here.
In the cloud era, Taobao launched a very "cloud" characteristics of the customer service model - cloud customer service, cloud customer service to the community like to help people and have the ability to help people Taobao people together, so that customer service personnel at home or at school to provide remote service to customers. Provide remote service, realize the "HO (Home Office, home office)", and make full use of the customer service staff's scattered time, not only reduces the cost, but also improves the efficiency. Of course, this "cloud" is not the true meaning of cloud technology, just a concept and gimmick. And in the era of big data, what business value will be brought to customer service?
I believe that big data will bring a change to customer service, bringing great imagination and unlimited prospects for development of customer service. It can even enable the customer service department to transform from the original cost center (high cost, low value) into a profit center (to enhance brand value and create revenue). Here I take three examples to discuss with you and look forward to the future of customer service.
One, intelligent voice customer service
At present, communication carriers and other more advanced enterprises in the field of customer service has realized intelligent text customer service, through text recognition technology and intelligent matching algorithms through SMS and website text customer service service service requests intelligently match the answer, without human judgment. To realize intelligent voice customer service, we also need to pass the two hurdles of recognition and matching.
Let's start with recognition. As early as Siri before, there have been a lot of voice recognition tools have come out, the earliest electronic computer-based voice recognition system is developed by AT&T Bell Labs Audrey voice recognition system, which can recognize 10 English numbers, and now AT&T's voice system Watson has been able to realize the real-time interpretation of online German and English. With the current technology, speech recognition is still difficult, the main difficulties faced by 2:
1. algorithm
Algorithm is the core of the software, the current speech recognition algorithm using the language model is still a probabilistic model, has not yet been developed into a linguistic-based grammar model, algorithms do not breakthroughs, the effect of the progress can not be achieved by leaps and bounds. The optimization of the algorithm is not an overnight thing, it needs to be done slowly and continuously, especially speech this kind of unstructured data (inconvenient to use the database two-dimensional logical table to represent the data), but with the development of big data analysis technology (for the management and analysis of unstructured data), it will also be a boon to the development of new algorithms. Some of the core algorithms such as feature extraction, search algorithms and adaptive algorithms are also improving step by step, and with the continuous enrichment of data sources, the algorithms are more and more accurate recognition.
2. Adaptability
Speech recognition algorithms are limited by factors such as dialect, tone of voice, environment, and timbre, which requires a certain degree of adaptability of the language recognition system, the recognition of different accents and dialects needs to be based on a large speech database, and the management and analysis of these unstructured data counts more on big data technology. As for the exclusion of environmental noise, timbre and other factors, I personally feel that we have to rely on advances in semiconductor sensing technology, and leave it to experts in the field of hardware to further explore.
The next step is to talk about matching. The matching algorithms are now relatively mature, perhaps not directly related to big data technology, but their accuracy depends on the richness of the data sources, and the need to dynamically adjust the matching results in the constantly generated "interactive data".
In summary, with more and more data sources, big data technology continues to progress, speech recognition system is also continuing to improve, in the end, the algorithm is still the core, while the data is the basis for this kind of unstructured data, perhaps the traditional database technology Handle can not be, but the big data technology is very promising. I believe that soon, voice recognition technology breakthroughs can not only realize the intelligent voice customer service, but also will change the way of interaction between people and things.
Two, voice text conversion
Because the core of this function is also voice recognition, so big data technology on the conversion of the accuracy of the guarantee support need not be said again. The reason why it is singled out to talk about it, because it has a role to play in customer service.
For the call center, customer service personnel and user calls are to record backup, these voice data can be really not small Oh, Guangdong Mobile, for example, Guangdong Mobile Customer Service Center will be added each year about 60T of data storage, this volume for the general business has been "big data". It is reported that these data are used to save the tape, and these should be saved for decades can not be destroyed, think of the time when the light of these tapes occupied by the room rent is a lot of money ah, not to mention other costs. And if these speech can be accurately converted into text, text storage takes up much less space (a mobile hard disk can store a library of data volume), storage costs are simply plummeting, not only to achieve cost-effective, but also for the natural environment is also a kind of good.
Some people may question whether these recordings are intended to be traceable, not the original audio recordings, and what if the customer doesn't recognize them? Of course, I have to declare that not all recordings should be converted into text, for customer complaints or business calls, still retain the audio recordings, on the one hand, it is easy for enterprises to customer service attitude (speaking tone of voice or something really rely on the language to be able to judge) and the quality of sampling, and on the other hand, back up to stay in evidence. For more consultation or query calls, usually do not have to leave evidence, the voice into text, not only reduces the storage space, these text data can also be used for subsequent information mining, used to improve service or find business opportunities, after all, the text of the information analysis is much easier than the voice.
Three, customer information mining
In the Internet era, in addition to the number of users, turnover, etc., data has been recognized as the core resource of the future. I remember Jack Ma once said something like this: "Do you know which province has the largest woman's bust? Do you know which city's men like to use what brand of clothes, perfume? You do not know, Taobao know." How many companies pay attention to the Taobao User Behavior Report every year in an attempt to dig out some data to boost their sales, and from here, the value of data is evident.
And the customer service department as the front-end of the enterprise customer direct contact window, every day from the customer can get a lot of information, and even in the customer is more satisfied, take the initiative to get some hobbies, occupations, and other information, accumulation of small amounts into a large number of certain times, these data will be for the enterprise huge value. Of course, the entry of these data can not rely only on manual, which involves more customer view and labeling issues, to be analyzed below to think about. The application of customer information mining, I would like to give two simple examples. For example, through data mining, it is possible to find out which users are golf enthusiasts and conduct accurate marketing to avoid customer resentment and complaints caused by blind marketing. Another example, with the positioning technology has become the standard equipment of cell phones, personal location information has become a gold mine to be mined in the field of customer service, foreign carriers have begun to analyze the data of these personal location information, and will provide insights for government and enterprise customers, these location information can provide the basis for the enterprise's physical stores, business halls to choose the location.
The above is what I share with you about the big data era, how customer service will be changed related content, more information can be concerned about the Global Ivy share more dry goods