Questions on Text Analytics

Text analytics is a way to extract information from unstructured and semi-structured documents and texts that gives you a more complete view of your business, your customers and competitors.  This method deals with huge bulks of texts and textual data for the purpose of revealing the intentions and sentiment of customers.

Text analytics provides crucial insights that drive enterprise decisions and helps businesses improve.

It can be and is widely employed in analysing customer feedback, tweets, blogs, call centres, newspaper articles, email, web pages and many more.

Text analytics leverages the latest technology and knowledge in linguistics, IT, science and AI.

Some of the unique uses of text analytics:

  • Customer Feedback – Analyse unsolicited feedback left by customers via call centres, email or contact forms.
  • Open Ended Survey Questions – Get your fingers on insights from answers to open-ended survey questions
  • Ratings and Reviews – Tap into the endless stream of insights from ratings and reviews on your own platform or on sites such as Amazon, Yelp! and many others
  • Social Media – Gauge the trends in posts, comments, tweets and replies on social media

Technology applied in text analytics

Technology applied in text analytics could be AI, Machine Learning, Boolean rules and tailored sentiment models that bring about great precision specific to various industries.

The end goal of any Text Analytics exercise is to extract insights from structured or unstructured text. In general, there are 3 pieces of information that we can analyze with TA software, including TalkAboutIT.

  • Sentiment – Or more specifically the emotion shown in the text. This is done via algorithm without human intervention. In TalkAboutIT we use a Sentiment algorithm.

 

  • Categories/Topics – Or what the person is talking about. Is the user complaining about the Internet Speed or the Price of the Internet and etc. We are extracting this information with a mix between human and automatic coding. A Text Analyst creates a set of comprehensive Boolean rules that are designed to identify specific themes.

 

The rules are tested on a small Training dataset with 20-30,000 comments, and once the Code Frame has been finalized, the rules are applied to the full set of comments (hundreds of thousands).

 

  • Metadata – This software incorporates additional data that has been supplied together with the source text into its dashboards. For example, we can combine the existing NPS score of each comment that is entered into the system with the Sentiment and Category extracted from TAIT and visualize them on a single Dashboard.

 

Advanced features of TalkAboutIT

However, the ability to quickly crunch and analyze the text is not the only strength of TalkAboutIT. While developing the tool we always had advanced users in mind as well, and we added features that, while not mandatory for every task, will be of great value to CEM who would like to take Text Analytics to the next level.

 

  • Sentiment override: the function allows the user to change the value of the Sentiment score of comments captured by specific rules. Sometimes, the question of the survey is sentiment bearing, and we need to account for that. For example:

 

Both questions have the same answer, but one of them is very positive the other very Negative. With TalkAboutIT you can use rules and metadata filters to ensure this text is coded accurately.

  • Metadata filters: Allows the user to write rules that work only if the correct Metadata is pertinent to the text. For example, in the Topic for Internet speed you will specifically exclude all comments that are from Mobile Internet Surveys/Questions to avoid any false positives.

 

  • User specified exports: TalkAboutIT has advanced export functionality that allows the user to select what data to be exported. The system can export both the original data that was imported in the system, but can also export the raw outputs from the Text Analytics. The outputs include the texts, relevant metadata, the relevant categories, and the sentiment. Exporting the output allows the user to modify or deep-dive in the data using alternative software for statistical and data analysis. Moreover, GemSeek has extensive experience with Data Analytics and can assist the client in timely and efficient manner.

 

TalkAboutIT as an enterprise solution

The ability to quickly analyze small samples of texts is not the only way to use our tool. The modular approach taken during the development of TalkAboutIT allows it to function both as a SaaS and as Enterprise solution that can be integrated with yours or other systems to deliver analysis automatically in real time.

  • Using our API integration, you can send unstructured and structured texts to TalkAboutIT which will process them and send them back via the API to your system. Once the connection, and Code Frame (list of Categories and associated rules) are developed, the process works without human intervention and with only marginal delay of a few minutes required by the processing time.

 

  • TalkAboutIT can seamless be integrated with RecommendIT 3.0, its sister full service customer experience management platform. This will allow your business to close the loop from the first creation of the questioner, to the live dashboards visualizing the processed results

Text analytics professionals

Companies that offer text analytics solutions usually employ the super powers of customer experience professionals, data scientists, IT and linguistic experts.

Text analytics methods and practices

Text analytics methods and practices include a lot of analytical activities such as sentiment analysis, stigmatization, visualization, text identification, text mining, text categorization, text clustering, search access, entity modelling, and link analysis.

What is sentiment analysis? This process identifies and categorizes an expressed opinion by a person that is in a written form of text. It determines the attitude of that person towards a specific topic, service or a product. It tells you whether that particular person has a positive, negative or neutral attitude.

benefits from text analytics

You can easily identify the benefits from text analytics as well as the trends and insights that are important for your business or your organization. The benefits from text analytics can help you realize the issues and pinpoint the factors that make people or customers happy. Thanks to this analytical method you can unlock the voice of your customers and leverage it to improve your business.

Text analytics surveys

Text analytics surveys can extract some valuable information for an industry or particular business. A good example could be a survey with open-ended questions asking your customers about their experience with your products and services and why they would (not) recommend them to friends and relatives. This is a really effective way to collect customer or clients’ opinions and get a feel of how they look on your brand, product or service. And, on top of that you can get some suggestions for unique and innovative ideas.

text analytics matters

Text analytics matters a lot especially in today’s fast-paced world where the expectations of your clients continuously rise. As a result, you often have to run to simply stand still, and you need to be even better and faster to shine in terms of customer experience. You have to monitor closely the behaviour and experience of your clients and with the power of text analytics you can make more informed decisions on how to improve your business so that it better caters to your clients’ expectations.

Is text analytics for me? Is TalkAboutIT for me?

Do you sit on large volumes of free-form text that you would like to analyse? If yes, then TalkAboutIT is for you.

Text Analytics can be applied in virtually any industry. FMCG, Retail, Healthcare and Pharmaceuticals, Energy, Technology, Logistics, Telecommunications, Construction, Travel, Leisure, Hospitality, Professional Services, Consumer Durables, Capital goods, Government etc.

What does “ready-to-fire” mean?

“Ready-to-fire” is our vision for text analytics. You simply upload your data and the tool provides you accurate sentiment and topic categorization. However, none of the tools on the market can do that out-of-the-box, regardless of what they promise. In order to have a high accuracy and actionability of the analysis, you need to invest weeks into setting up and testing categorization rules that match your business specifics. Let’s be honest – no one likes this and a lot of people don’t really know how to do it. So our text analysts can do that for you. We make TalkAboutIT work for you and then it really is “ready-to-fire” – you can login, upload your data and have accurate insights within seconds.

Text analytics application can be found in many different industries such as the automotive one, the FMCG industry and industries that deal with luxury goods.

Some other areas where text analytics comes to good use include industries like Healthcare and Pharmaceuticals, the Energy industry, Technology, Logistics, Telecommunications, Construction and Lighting,  Travel, Leisure, Hospitality, Professional Services, Consumer Durables, Capital goods, Government etc.