4. HIPÓTESIS
6.3. Métodos de control de Diaphorina citri
6.3.1. Control biológico
Besides these, the TM Forum has also published the Big Data Analytics Guidebook (TM Forum, Big Data Analytics Guidebook) in which many use cases are included, mostly concerning BSS, but also including OSS scenarios for network fault prediction, for example.
5.5.1.1
A call to a telco Customer Care centre
Scenario Name A call to a telco Customer Care Centre
Background / Rationale
This scenario represents a future daily situation in which Big Data technology can bring telecom operators and customers together. The most interesting side of this story is how far this is from reality today in the Customer Care domain.
Scenario description
A customer is using his smartphone when suddenly a message appears on the screen indicating “Emergency Calls Only” and the call is suddenly
(Storyboard) dropped. He checks the rest of applications that were running in parallel. He had a VoIP application with several on-going conversations that seems to be disconnected now. He was also uploading a video to his favourite social network, which has not been totally completed according to the status shown on the screen. Finally, he was in the middle of a process for downloading some MP3 files from an on-line store who has already received the payment for the merchandise but not all the files have been successfully retrieved.
He resets his smartphone and checks that ordinary telephony works but not his internet connection. He decides to call the Customer Care (CC) number and is immediately served by a friendly agent who knows his name. When he begins reporting his problem, the customer care agent (CSA) listens carefully and registers the description of the incident from the customer’s perspective in written text. The CSA says that optionally he can attach the voice conversation to the customer’s file, which he does. He asks the customer whether he had open applications when the incident occurred, out of a list of the apps the customer has installed in his smartphone, which is known at the CC.
The CCA checks the network nodes status in the incidental area and immediately sees that there has been a problem in one of them. He sends an instant notification to the technical support team, who are already aware of the incident and who report that the problem will be fixed in 5 minutes.
The CCA checks additional information concerning this particular customer. Not only he has been a customer for more that 10 years, but he has recently subscribed to one of their ADSL offerings. The operator has punctually received payments for the unified electronic bills available in the system. He has also raised several complaints in the past concerning the quality of his VoIP service in multi-party business calls when a number of users are involved. He also gathers his most used numbers for calls and for SMS.
The usage information concerning the uplink and downlink transactions with unachieved service delivery are disregarded for billing by the CSA. This action of discarding the records is registered in the system for future uses, as well as the whole procedure. The CSA informs the customer that he can relaunch both transactions with no duplicated cost. The payment for the music records will be cancelled with the content provider by the CSA.
Besides, the CSA considers that this customer deserves a loyalty action to be applied, the CCA proposes to activate two separate discounts on calls and SMS to those favourite numbers that will begin being applied immediately.
Right afterwards, the customer writes an entry in his blog explaining how well he has been served by his telecom provider.
Functional areas covered by the scenario
Network:
Reduce Problem Resolution Time Increase Staff Efficiency
CDR analysis
Sales / Marketing /Profitability Manage Vendor Performance Advanced Customer Segmentation Product/Service immediate Offering Foresee / Reduce Customer Churn Gain Insights on Customer Behaviour Convergent & intelligent Customer Care
Technical domains involved
Data acquisition, data analysis, data curation, data storage, data usage
5.5.1.2
Advanced customer segmentation
Scenario Name Advanced customer segmentation
Background / Rationale
This scenario is partially based on a use case carried out by British Telecom and reported in the Global Information Technology Report 2012 by the World Economic Forum [WEF01].
Scenario description (Storyboard)
This use case shows how the combination of data coming from different sources (web, geographical data, network information) can be used as an input for advanced customer segmentation and sales regions definition which can, in turn, help sales departments offering strategies.
A hundred million links among more than 20 million numbers made anonymous from an original database of some 8 billion telephones can be examined and the analysis of the resulting network to identify natural communities in the data, (where a community is characterised by relatively dense within-group links and proportionally fewer out-group connections.) This information can be combined with information retrieved from social media. What are these customers saying in their favourite social networks to one another? What are they being told? What is being said about a new product? How often is it mentioned and how far does it get? What kind of blogs does my product appear on? Are they specialised, generic, complaints forums, etc.?
Another authorised survey of thousand households and the association of more than a million call records to their responses made possible to assess the classification of households according to their calling networks. The results suggest that some dimensions of social interaction can serve as reasonable predictors of whether a household is comprised of “Alone, over 56,” “Couple, both aged over 55 with no cohabiting children,” or ”Couple, with children aged under 12”.
A community is almost synonymous with "segment." A segment is a group of customers that will react similarly to a message. This helps identify new marketing campaigns and defining business opportunities and the right moment to communicate with customers. Communities can be found in many contexts, but broad criteria from a telecom perspective might include shared calling groups, locality, age, interests, language and subcultures. Telecom operators would seem to have particular advantages in regard to community-based marketing. They have unprecedented, aggregate detail about customers’ communications habits and movements as well as a feedback channel that could be used for tailored messages.
Functional areas covered by the scenario Web: Sentiment analysis Network: Traffic analysis
Advanced Customer Segmentation Product/Service Offering customisation Foresee / Reduce Customer Churn Gain Insights on Customer Behaviour Technical
domains involved
Data acquisition, data analysis, data curation, data storage, data usage
5.5.1.3
Telecom customer journey
Scenario Name Telecom customer journey
Background / Rationale
This scenario tries to go through the customer journey in order to understand what data is exposed and how does the customer experience the relationship of service and trust with the telecom operator.
Scenario description (Storyboard)
A telecom customer sees a product especially pushed for him on a website. He buys and pays online via laptop. The customer is notified of delivery time on his personal cell phone and decides to collect the item directly from the point of sales.
He registers via tablet to receive further notifications and makes several calls to the customer service because there is a missing accessory in the box. The separate piece is delivered separately.
Later one, the customer tweets about it and uploads photos of his new device via Facebook, which is liked by his circle of friends. They are all planning a trip to Indonesia according to the comments on their walls. They are also fans of cultural events and share this sort of information in their community.
Before her departure, the operator can offer a special international roaming plan for calls in Indonesia. Besides, the group of friends might be offered a holiday package with local lodging companies. The operator can also offer a notification service about local festivals and cultural events when abroad. Functional areas covered by the scenario Web: Sentiment analysis Network: Traffic analysis
Sales / Marketing /Profitability Advanced Customer Segmentation Product/Service Offering customisation Foresee / Reduce Customer Churn Gain Insights on Customer Behaviour Technical
domains involved
5.5.1.4
Dynamic bandwidth increase
Scenario Name Dynamic bandwidth increase
Background / Rationale
This example illustrates how the information retrieved by Call Centres can help identify infrastructure and network problems
Scenario description (Storyboard)
In most organisations, the customer care data is analysed typically from a SLA(Service Level Agreement) perspective. For example, turnaround time, average wait time, etc. are often measured and ensured. However, a greater insight can be gained by, e.g. the actual transcript of the conversation. This could even lead to the identification of problems regarding the telecom infrastructure (e.g. infrastructure bottlenecks).
Telecom providers are currently getting more revenue from data services than from voice services. This is why operators are very interested in launching new services that generate a lot of traffic, such as cloud-based gaming, for example.
In this competitive environment a telecom provider launches a new viral gaming application on mobile devices. A few days after its launch the operator observes a burst of calls to the call centres and on text mining the transcript data specialists find a great increase in the keywords alluding to performance. The specific intelligence regarding keyword burst and specific time of day at which this was encountered can derive an automatic actuation on the network in order to dynamically change the provided the bandwidth based on usage.
Functional areas covered by the scenario Web: Sentiment analysis Network:
Network service enhancement Sales / Marketing /Profitability Customer satisfaction
Technical domains involved
Data acquisition, data analysis, data curation, data storage, data usage
5.5.1.5
Security application based on cell towers
Scenario Name Security application based on cell towers
Background / Rationale
This example illustrates how the information retrieved by Call Centres can help identify infrastructure and network problems
Scenario description (Storyboard)
When a call is made, the operator usually captures data such as the subscriber, the time and the duration. Depending on the type of call and service used, additional data can be gathered. For example, serving switch data, serving cell tower IDs, device identification (serial) numbers, as well as International Mobile Subscriber Identity (IMSI) and International Mobile Equipment Identity (IMEI) codes. The unique ID of the cell tower a handset was connected to when a connection was made can be used for collocation analysis.
By examining terabytes of CDR/Tower records from the switch it is possible to triangulate on a few collocation events. A co-location event can be defined as the same mobile tower being used to route calls during a specific point in time. The combination of massive Hadoop clusters and columnar database architecture allows these queries to be executed at great speed in order to retrieve a reduced set of records to analyse further.
This allows to identify if the same person has used several devices by combining CDR information (IMEIs or IMSIs) with network information (registered by the mobile tower).
Functional areas covered by the scenario
Network:
CDR/tower information
Sales / Marketing /Profitability Security application
Technical domains involved
Data acquisition, data analysis, data curation, data storage, data usage