We stand with Ukraine

Text Analytics best practices and applications

Expert.ai Team - 25 October 2016

With many years of experience in developing and implementing different kinds of information management or process automation process projects, we are in a unique position to share what we believe are the most effective text analytics best practices and the most innovative and value rich text analytics applications.

Let’s start with Text Analytics best practices by focusing on a couple of areas that can have a significant impact on a project’s success, but that are often ignored by customers and/or consultants.

Text Analytics best practices

  1. In Big Data text analytics applications like innovation or security intelligence projects, you don’t really need to be selective in choosing the sources or content to include in the analysis. One of the biggest advantages of working with an advanced text analytics engine is that it does a lot of the heavy lifting for you! The engine can analyze content at scale, and it offers plenty of capabilities and criteria for intelligently filtering the content. This approach delivers results that are both manageable and actionable because you can accurately eliminate all of the irrelevant content in the post-acquisition phase. Given that extremely relevant data are often retrieved from the most unlikely, unexpected sources, this text analytics best practice is essential because it allows for more effective results and higher ROI.
  2. If you have advanced needs for information intelligence or process automation processes, make sure you include a semantic based engine in the text analytics application selection process. In this way, you can compare and contrast the results with the results of pure statistics or machine learning based engines. A semantic based gives you the ability to understand the meaning of words in context from the beginning. These systems still need a bit of refinement to augment the knowledge with domain- or company-specific knowledge, but they do not require the cumbersome and resource-intensive training that pure machine learning systems require. While this text analytics best practice might require disclosure that our Cogito platform uses this exact approach, our many years of experience working on a wide variety of advanced projects supports its validity as a best practice. In the real world, for text analytics applications, the pure machine learning approach is simply too long and too expensive. Achieving significant ROI in a reasonable timeframe with such a heavy investment is very difficult.

Text Analytics applications

Together with the explosion of available information we are all experiencing, the number of text analytics applications that provide business value to organizations are growing by the day. The three applications below, while perhaps less known, offer organizations the highest ROI potential, based on our experience in the field:

  1. Automating the insurance claim process. This is a very expensive process for any organization in the sector. Outsourcing claims to external providers (as many companies have done) is only a partial solution. Often, the negative impact on customer satisfaction resulting from inexperienced employees working on claims offsets any initial savings. With an advanced engine, this activity can be automated with an text analytics application that automatically extracts data from all the forms submitted by customers, compares them with customer coverage data in the company DB, and recommends approve/reject to the agent in charge of the final decision. In this scenario, you can expect savings of 50-60% when compared to outsourcing, with less exposure to errors.
  2. Automating customer interactions. This text analytics application includes self-service apps accessible from a website or via text message, as well as the automatic categorization of emails or support to call center agents through natural language based question answering engines. The utility and popularity of intelligent smartphone agents is increasing both awareness and adoption of these types of text analytics applications in many types of organizations.
  3. Intelligence event tracking applications. There is a growing importance for organizations to be able to discover, anticipate or just tracking events that impact their various activities. The cost of a partial understanding of events around us can be catastrophic in certain scenarios. Just imagine the impact of not knowing what the competition is doing; if product development is behind the curve in learning about new market trends and disruptive technologies; or the operational risks of trusting the wrong supply chain partners. Companies are surprisingly still using very manual processes to track these situations, often resulting in learning too late about a critical event. An advanced text analytics application, through fast and accurate understanding of information streams and content, can be your eyes and ears, your sensors in the world, working 24/7 to track what is happening in the echo-system, isolating events and alerting you in real time.