Nowadays, people rely on search engines and chatbots to retrieve and share information from various
resources. Search can be carried out mainly in two ways, keyword search and semantic search. In a simple
search, searching is carried out word to word, while in the keyword search, the upper case, the lower case,
and order is ignored, and the specific results having the exact words are returned. Semantic search is a
searching technique where the meaning of the searched query is being looked for, along with its context,
so that the most relevant searches can be obtained with maximum information. Unlike the keyword
search, which provides the exact word results without knowing its context and displaying limited insights
to the users/customers. The search engines do not currently offer accurate search with the best relevance
as it is based on keyword search and not semantic search. Public To put this into perspective, if the
customer/user searches, "what is my destination pickup point? "Keyword search would try to return the
sentences having the exact same words from the dataset, ignoring the sequence and case sensitiveness.
While if the same input is fed to the search engine using the semantic search technique, it will return all
the results with a similar meaning like "last stop", "address to drop", etc., and the keyword search. This
gives the user/customer a more accurate and broader scope of insights for the searched query. As
ambiguous data is being obtained from other search techniques, that makes it difficult for the users to get
the desired results. These ambiguities from the results are reduced using the semantic search approach
by filtering out the results that have less relevance with respect to the context of the query. The broad
range of my project includes Artificial intelligence, Semantic search, and Natural Language Processing
techniques to give the highest relevant search result based on domain knowledge (keywords).
The focus of the current project is on the following areas:
● NATURAL LANGUAGE PROCESSING: To extract the correct information.
● SEMANTIC SEARCH: To find the relation between search and the data.
● INTEGRATION: To develop a CHATBOT for interactive sessions.
Hosting the developed chatbot on the cloud saves cost, makes it more secure, and achieves the highest
availability. AWS (Amazon Web Services) is used as the cloud provider.
Business Impact:
• Customers face problems with some off-highway machine equipment’s and try to reach out to the
concerned team.
• Every time raising a concern, if the problem is already solved through raising request to application
team the same solution can be provided to the customer.
• This saves manual effort, time, and cost for both customers and employees.
• An accurate search system with user friendly UI like Chatbot would be helpful for customers ease
of use