How agencies can help constituents find accurate information quicker
COMMENTARY | With AI-enabled answer engines, users can ask complete questions and get the correct answer regardless of how they phrase the question.
In today's world of search, people expect immediate answers to their questions. Yet, despite their best efforts to provide timely responses and the significant investments already made, many state and local agencies are struggling to quickly and accurately respond to residents' questions. This is primarily due to poorly performing keyword search engines and chatbots as well as understaffed call centers that must deal with all the inquiries that could not be answered through digital channels.
The underlying cause for this situation is that 80% of enterprise data is unstructured, i.e. free-form text documents, which makes it extremely difficult for traditional search tools and chatbots that use keyword- or token-based methods to identify and retrieve the information needed. These rule-based approaches may fall short on answering user questions.
State and local government agencies have tried several approaches to address the challenges of unstructured data, including:
- Adding structure to unstructured data with taxonomies and ontologies.
- Leveraging large language models and deep learning.
- Implementing symbolic artificial intelligence that embeds human knowledge and
behavior rules into computer programs.
However, to date, none of these approaches have substantially improved citizens’ ability to quickly find answers. Agencies need an “answer engine” that can surface relevant information instantly.
AI-enabled answer engines enable users to ask complete questions as if they were asking a subject matter expert and get the correct answer regardless of how they phrase the question. They also allow business users and subject matter experts to leverage AI technologies directly without relying on technical teams for support.
Under the hood, the inherent capabilities of these answer engines include important features such as:
- Natural language understanding: A broad set of cognitive techniques that understand the nuances of language, including domain-specific terms.
- Query resiliency: Consistent results no matter how multiple individual users phrase the question.
- User intent: Understanding the intent of words and phrases. For example, it understands that a user asking questions starting with who, when, where, etc. may be looking for data on a specific name, location, timeframe or measure rather than a direct token match.
- Multi-level explainability: Contrary to most “black-box” AI tools, answer engines provide users with the ability to see the various strategies and scores used to find each result, boosting users’ confidence while also allowing them to tune the platform to incorporate new products, focus areas and terminology.
- Easy usability and manageability: Answer engines use no-code platforms that can eliminate the need for labeling/annotating or model training.
The private sector is already transitioning to these state-of-the-art, AI-powered answer engines to help customers locate the right information when they need it. In time, this trend will make its way into government.
All levels of state and local government must address how they can be more responsive to citizens' questions. While time-to-answer is always a key consideration to driving user satisfaction, it is the potential of eroding trust with wrong answers and inadvertently denying citizens fast access to key resources that should be driving government agencies to answer citizens' questions more quickly.
A word about ChatGPT…
ChatGPT has shown the world what is possible in search. Although it has made great strides and could be suitable for many consumer search applications, ChatGPT is not quite an answer engine because it too often returns unacceptable levels of inaccurate information and does not provide any explainability as to how it generated a response. Government agencies who must provide both trusted answers and transparency cannot afford to rely on tools like ChatGPT … just yet anyway.
Today, agencies understand the importance of delivering timely and accurate answers and the impact that service has on citizen experience, maintaining trust in government, and ensuring that marginalized citizens are not further disadvantaged by long wait times. Despite the availability of new technologies and approaches like answer engines, the reality is that fully implementing the changes that are necessary to achieve these goals will take time and a sustained effort from agencies and government leaders. Nevertheless, the benefits to all will be well worth the effort.
John Reuter is the chief strategy officer at Kyndi.