– Contextual search is a powerful way to gather personalized results based on information from multiple structured and unstructured data repositories.
– Without context, users often must sift through a bunch of irrelevant results before finding what they want.
– Contextual intelligence helps to assign confidence rankings to search results and streamline the process of finding relevant, current data.
If you are a hunter, a soldier or a police officer, a bullet point is the tip of a ballistic projectile. Yet if you are a writer, marketer or editor, a bullet point is simply an item in an unstructured list.
This is a simple example of how contextual search is a powerful way to gather personalized results based on information from multiple structured and unstructured data repositories. Further, it shows how different people use the same keywords or phrases to seek out unique content. Without the benefit of context, users often must sift through a bunch of irrelevant results before finding what they want.
Even within an enterprise, departmental users have unique corporate data perspectives based on their roles and responsibilities. Role-based permissions are fairly common. Applying additional contextual intelligence to search results, and according to a recent Findwise survey, improving discoverability generated:
Increased knowledge sharing and retention
Greater re-use of content
Improved decision support
Integration of isolated content repositories
Greater eDiscovery/Compliance effectiveness
Contextual intelligence helps to assign confidence rankings to search results and streamline the process of finding relevant, current data.
Conventional search engines index internal and external content and data to the extent they are able, and every search query will solely be based on the literal search phrase. Vast amounts of new data and content are constantly generated, and the ability to find accurate, timely information using traditional methods isn’t keeping pace. A search engine built on cognitive technology ingests and indexes content, and as it learns about the user’s role and demographic information, it will apply “filters” to the search results, to prioritize and rank the results to produce relevant data. It may prioritize a list of data sources or URLs based on different factors.
The physical proximity of the user to a customer, supplier or business partner can prioritize a search result ranking. The geographical location might also have cultural or infrastructure factors that need to be taken into account. Algorithms can be programmed for regional preferences related to language or religious characteristics to rank search results. There might be a great retail location under ten miles away from a user, but if it is across a lake, or if IoT road network sensors detect a traffic jam, other locations may be more desirable.
Search or result preference history
Contextual search functions built on cognitive technology learn as time passes, so as a user conducts multiple searches over time and selects certain results, personalized result sets can be delivered and prioritized.
User attributes like age, gender, income and education all have significant influences on what we look for. If a search bot can return a set of personalized options based on popular selections from other users with similar profiles, it makes for a better user or customer experience. For B2C e-commerce websites like Homebase with large product catalogs, or data-intensive loyalty programs like Boots UK, cognitive and contextual search technology enables personalized search results and marketing content that’s tailored to the user. It has a higher propensity to convert browsers to buyers and increase customer loyalty.
Natural language search terms
Traditional and Boolean search terms used by search engines can be limiting. Cognitive technology empowers users to search for information using conversational language, like “Where can I buy a domestic hybrid sedan locally for under $40,000?”. The contextual search engine could apply localization intelligence to the user to determine dealers in their area which sell hybrid vehicles, the user may have some search history which demonstrates a preference for a certain domestic brand of vehicle, and prioritize search results for private sellers based on search history as well based on predictive modeling.
Virtual shopping assistant chatbots can take users through a conversation process to better define their needs. Query calls to a context-enabled search mechanism can make a user feel like they’re chatting with a real customer agent, and improve online buying experience. Similar scenarios help researchers find the latest data for reports, or for colleges and universities to respond to student questions.
Searching across multiple stores of data and content
APIs ike Watson Discovery Service can be connected to structured and unstructured information sources like:
Public domain government and NGO research websites
Social media, news and consumer review channels including audio, video and text content
Internet, intranet and extranet portals on the semantic web
Data such as weather or equipment maintenance requirements from sensors connected to the Internet of Things
Cloud-based and on-premise database-driven applications
Document management and business process management systems
Gathering insights from these repositories is like finding a needle in an exponentially expanding haystack. Context helps to isolate results and rank their likelihood of addressing the query, without the user having to conduct multiple searches. User and business case relevance improves, and users still have the option of investigating multiple sources based on confidence ranking.
Context enriches the value of data, wherever it resides, and defines whether data is contextually rewarding to business professionals and consumers, or unlikely to address their needs. It creates a corpus of data for users to extract insights from, instead of having to execute multiple searches through zettabytes of structured and unstructured data.
Surface answers and patterns faster with Watson Discovery Service
The IBM Watson Discovery Service enables developers to work with APIs to enable contextual searches across structured and structured data and content. For further information about Watson Discovery Services, or to access a 30-day trial, please visit our website or contact the IBM Bluemix team.
See how you can use Discovery Service to quickly surface business critical-insights.
Natural Language Processing (NLP) is a massive space within artificial intelligence (AI), and enterprises are currently integrating NLP technologies into their existing platforms more every day. ...read more