Menu
Gergely Havlicsek, Product Owner, g.havlicsek@mito.hu
August, 2020 · 5 min read
Reacting to
certain
keywords
?
1on1
personalized
messages
To be able to understand users contextually
Providing relevant, easily searched answers for a chatbot
NLP is a process where we recognize intents (verbs, activities that the user wants to do) and entities (content for the action that needs to be performed) from the user inputs, chat messages.
So structuring unstructured data and giving back useful information to the user, requires a systematic approach in each step of the process.
Intentions of the end-user, which are conveyed by the user to your bot.
1. Casual Intents
‘Hi’ or ‘Thanks for talking to me’
2. Business Intents
‘What are your opening times?’ or ‘Is parking free?’
Use real discussions to train your bot besides manufactured and generated samples
Based on intent recognition the tone of voice and the flow is selected
Knowledge base integration with chatbot systems can be done on different levels, determined by the needs and use cases. Starting with simple search queries for relevant data you can advance the capabilities of your knowledge base integration to reach higher levels.
Eventually combining the advantages of NLP and Knowledge Bases.
1. Properly functioning search system
2. Stable and effective backend storage system
3. Simple archive and categorization system
4. Good integration capabilities
5. Analytics and reports
We focus on topics suitable for NLP which then can be scaled with the help of a knowledge base.
Our own methodology of NLP training follows a step-by-step cyclical process which is similar to agile programming. We put strong emphasis on creating a stable working model at the start and improving on it throughout the life of the project.
Every step has its own challenges and best practices, and we believe that a systematic approach can help us deliver the needed expertise at every step of the journey
Our own methodology of NLP training follows a step-by-step cyclical process which is similar to agile programming. We put strong emphasis on creating a stable working model at the start and improving on it throughout the life of the project.
Identify intents in advance
Differentiate between general/casual and business intents.
Train intents with original conversations
Otherwise train with manufactured utterances. Minimum 5, optimally 10 utterances.
Train - Converse - re-train
The feedback loop must continue in order to train your NLP models.
1-3 months
Setup
Identification of 5 most common topics, which are most suitable for NLP
3-6 months
Iteration
Testing and iterating on original scope based on real conversations
6-12 months
Feature development
Enough data is available to re-cluster topics and iterate feature development roadmap
Are you operating a large scale customer service with main clusters of recurring questions / topics in high volume?
How easy is it to integrate a new solution to your customer journeys and potential business backend systems?
Easy to start and collect feedback
Highly scalable solution
A huge step towards 1on1 personalization
You can find everything in one place here: mitobot.io
A kosárlabda csapatok 5 fős felállásban játszanak. A nevezett csapatokat csoportokba osztjuk, a sorsolást a tornát megelőző héten minden csapat részére kiküldjük. A mérkőzéseket hivatalos játékvezető vezeti. A mérkőzések egy pályán zajlanak. A csoportok első két helyezettje az egyenes kieséses szakaszba kvalifikálja magát. Az egyenes kiesés szakaszban a csoportmérkőzések során első helyezett csapatok a másik csoport második helyezettjével játszanak, ezt követően a vesztesek a harmadik helyért mérkőznek meg egymással, majd a győztesek vívják a döntőt.
Csapatlétszám: 5+1 fő.
Játékidő: 2×10 perc (szünet nélkül térfélcserével)
A mérkőzéseket 3×2 méteres kapukra játsszuk.
Page speed has become an important factor for website owners and SEOs since Google began focusing more and more on user experience factors. If companies don’t adapt to this new mindset, they will lose their organic power.
Nokia has experienced an extensive page load time on most of its pages, but lacked proper measurable information about the performance of the website. It is challenging to adapt site-wise technical modifications without understanding what is really behind the curtain.
Mito introduced a multifactor, multi locational-targeting page speed measurement and reporting system across different devices.
Our focus was to develop a measurement system from scratch that is able to report on different page speed KPIs so we can better understand what additional analyses are needed to increase the pages’ performance. We did not only focus on a page level measurement, but also grouped pages to page types that assists us in making more educated assumptions.
We paid attention to location modularity in order to extend the number of analysed locations to additional countries by adding new Virtual Machines to our framework.
We are able to measure different KPIs on both mobile and desktop so we can see how the site performance is realized across devices.
Our measurement system and dashboard provided us insights on what type of elements of the different page types might cause low speed performance and it helped us to further analyse the loading curve, just as the opportunities in Google PageSpeed Insights recommendations for the different page types.
We set up the measurement system and the dashboard that reports on website performance and opportunities.
Through properly tracking the page speed improvements, Nokia becomes a more powerful organic participant in the telco market in Google search, while they provide a better user experience for their visitors.
An international business of this size requires a comprehensive measurement strategy including responsible data collection, consistent management of tags and measurement codes, personalized website tracking and purposeful conversion rate optimization (CRO).
We kicked off Analytics projects from January, 2020, beginning with extensive Google Tag Manager and Google Analytics audits. Besides inspecting the different system settings, we also optimized the measurement and pixel implementation processes within Nokia in order to increase efficiency and reduce implementation periods.
After the key adjustments and process harmonization, we have moved forward and started to work on the website and landing pages’ efficiency, including several conversion rate optimization tasks such as heatmap analysis and A/B testing.
Nokia is receiving continuous support from our analytics team to ensure the client’s measurement and data is reliable. Due to streamlined processes, implementation periods are shorter and we eliminated unnecessary website tracking codes which cause longer loading times.
Nokia launched a digital campaign to educate the market that their industrial-focused LTE solutions are the key in accelerating the ultimate benefits of digital transformation and realization of the fourth industrial revolution.
The goal: to build brand awareness and create demand for Nokia private Industrial Wireless solutions in new vertical markets, especially infrastructure or asset-heavy industries where digitization has been mostly restricted to point solutions without general network connectivity.
We’ve developed a multi-touch global digital media campaign, which covered the entire user journey of business leaders, from education to making a business decision.
Our main goal was to make Nokia visible on a large scale through automated solutions and smart targeting to reach people when they are reading, talking, learning about or searching for private wireless networks.
To archive this, we’ve utilized the combination of digital media platforms, including industry specific content sponsorships, social media ads, paid search, programmatic hyperlocal and Digital out-of-home ads in front of target company HQs. We’ve made sure that when our key audience is browsing the web, using social media on their phones or just walking outside of their offices, they see a relevant Nokia ad.
The campaign has reached over 80 000 business decision makers and influencers in 3500 target companies, across 5 industry verticals globally.
Private wireless solution specific brand search volume increased by 70% during the campaign period (YoY). In Q2 and Q3 2020, Nokia Enterprise had a double digit year-on-year growth in net sales and new customers.