Need help?

Type your question and our NLP technology will find an answer

Press "enter" to see your results


A.I. is the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand and model human intelligence, but artificial intelligence does not have to confine itself to methods that are biologically observable.
 
Efforts to make computers understand human languages, or natural languages, are described by a branch of artificial intelligence known as natural language processing (NLP); Inbenta is dedicated to developing NLP technologies and facilitating sophisticated, intelligent human-computer interactions.
 
Contact us for a live demo of how our AI-powered technology works.
Was this answer helpful?
Thanks!
Natural language is language that is developed and generated spontaneously by humans, using our innate ability to understand, process, and utilize language. Although natural languages have rules and structure, they are very different from artificially created languages (called constructed languages) like computer programming languages. People use natural languages (spoken, written, and signed) to communicate every day.
 
Because natural languages have not been 'designed' in the same way that formal languages are, they tend to have many ambiguities. The same word, phrase or even an entire sentence can have multiple meanings, and one concept may be expressed in multiple different ways. This means that natural language is very expressive, but also that there are opportunities for confusion and varied interpretations.
 
Watch our CEO Jordi Torras explain natural language.
 
Natural language processing, or NLP, is the field of AI that is concerned with programming computers to understand natural language.
 
Was this answer helpful?
Thanks!
SEO, or search engine optimization, is the name for a marketing strategy that uses a variety of techniques to improve the position of a web site in organic search results. The purpose of SEO is to get more visitors to a page, by ensuring it appears higher in search results and is visible to more people.
 
A website can be optimized for search by methods like editing content on the page to make it more relevant to common search keywords, or editing the coding of the page to make it easy to be indexed by a search engine.
 
Inbenta's Long-Tail SEO targets long-tail searches to drive targeted traffic to your site; long tail searches have 4-5 (or more) keywords and are more specific and less common than simple one- or two-word searches, and therefore have less SEO competition.
 
Contact us for a live demo of how our long-tail SEO can significantly boost your website traffic.
Was this answer helpful?
Thanks!

Machine learning is rooted in the study of pattern recognition and computational theory in artificial intelligence.

The technology uses an algorithm to teach the computer how to solve problems and gain insights from solving those problems. Machine learning optimizes the customer experience by using feedback loops to monitor and improve its predictions.

While humans would be overwhelmed with masses of data, machine learning thrives and is able to evolve its understanding to discover hidden opportunities.

With Inbenta Machine Learning your business will experience continual, marked improvement via evolving predictive models that deliver highly specific intelligent solutions. All of which are based on the ways your customers are behaving today, while also taking into account past experiences that let you examine how customers interact with you over time. All told it’s valuable insight into how to grow your business.

Watch our CEO, Jordi Torras, explain how machine learning combined with our patented NLP can help you!

Contact us for more information of how machine learning can help your business.

Was this answer helpful?
Thanks!

Semantic clustering helps your company discover gaps in your content to enrich your customer’s experience and reduce the number of unanswered questions.

Inbenta’s Semantic Clustering groups semantically equivalent search queries — words, phrases and sentences — into clusters based on meaning. The higher the number of questions, words and phrases with a similar meaning, the greater the cluster.

This gap analysis function allows our customers to discover where they need to add additional answers and information to make sure their customers can find what they want quickly and easily, without having to raise a support ticket.

Contact us for a live demo of how semantic clustering can increase customer satisfaction.

 
 
Was this answer helpful?
Thanks!
Ratings provide instant user feedback about the helpfulness of FAQs in your knowledge base. Users can give a simple 'Yes' or 'No' rating to each FAQ, and it's possible to add a text box for more detailed comments. Every detail will be tracked in Backstage, including what users asked (and when), which FAQs they clicked, and whether they felt their question was answered adequately. Individual users can be contacted directly about their rating and/or comments for more details.
 
Customers' ratings and comments provide valuable insight into the quality of knowledge base answers. A support center armed with this information can respond quickly to bad FAQs, re-word confusing or unclear answers to improve understandability, or add additional information to FAQs that aren't completely solving users' problems.
 
As the knowledge base improves over time, graphs in Backstage will track improvements in ratings to give a clear picture of the state of the knowledge base.
Was this answer helpful?
Thanks!
A.I. is the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand and model human intelligence, but artificial intelligence does not have to confine itself to methods that are biologically observable.
 
Efforts to make computers understand human languages, or natural languages, are described by a branch of artificial intelligence known as natural language processing (NLP); Inbenta is dedicated to developing NLP technologies and facilitating sophisticated, intelligent human-computer interactions.
 
Contact us for a live demo of how our AI-powered technology works.
Was this answer helpful?
Thanks!
Natural language is language that is developed and generated spontaneously by humans, using our innate ability to understand, process, and utilize language. Although natural languages have rules and structure, they are very different from artificially created languages (called constructed languages) like computer programming languages. People use natural languages (spoken, written, and signed) to communicate every day.
 
Because natural languages have not been 'designed' in the same way that formal languages are, they tend to have many ambiguities. The same word, phrase or even an entire sentence can have multiple meanings, and one concept may be expressed in multiple different ways. This means that natural language is very expressive, but also that there are opportunities for confusion and varied interpretations.
 
Watch our CEO Jordi Torras explain natural language.
 
Natural language processing, or NLP, is the field of AI that is concerned with programming computers to understand natural language.
 
Was this answer helpful?
Thanks!

Using dynamic FAQs, you can prevent the most common, repetitive support tickets from ever being created. Inbenta's software will analyze user questions as they are typed and provide helpful, informative answers from your knowledge base.

Dynamic FAQs make your support site responsive, allow it to adapt to your users from the first moment they arrive and giving you valuable knowledge about the questions your customers ask most.

 
Was this answer helpful?
Thanks!

From the moment your customer starts creating a support ticket, our intelligent instant answers suggests content related to their query to deflect unnecessary tickets.

Using our powerful semantic search engine, instant answers analyzes the user's input and directs them to helpful information, helping prevent your support agents from being inundated with repeated questions.

Instant answers lowers support center costs and boosts your self-service success by enabling customers to answer their own questions rather than sending an email or creating a support ticket.

 
 
Was this answer helpful?
Thanks!

People discuss your company and products every day online, and it's essential to be aware of current and potential customers' opinions.

Inbenta's Sentiment Analysis tool provides instant feedback on customers' feelings and attitudes towards your company in everything from social media posts to product reviews and customer support emails.

Our natural language processing technology automatically classifies each interaction as positive or negative in tone, offering global insight into your customer experience so you can always understand online opinions, market products effectively, and monitor your reputation.

For a live demo of how it works please get in touch!

Was this answer helpful?
Thanks!
Spelling correction is important both for content creators and for search accuracy.
 
Typos in knowledge base articles can confuse users or even give a poor impression of your brand, and misspelled words in search queries make it harder to detect meaning.
 
Our spell-checking tools can identify mistakes as editors create new content, and periodically provide a list of typos in existing FAQs so they can be corrected.
 
For search applications, our spelling corrector analyzes unknown words in user questions and applies corrections, allowing it to complete the search with the intended word.
Was this answer helpful?
Thanks!

Category trees will improve your support agent's response time by allowing you to store your content in specific categories which are easy to retrieve when agents are dealing with support tickets.

Once integrated, the category tree will show you your most popular content which means you can display content based on what’s most often clicked, so you know what the hottest topics are for your fans and communities. 

Contact us for a live demo of how category trees can enrich your customer experience.

Was this answer helpful?
Thanks!
A knowledge base is all the information your users may need. It may contain a list of FAQs, informative pages from your website, videos, PDFs, or other documents and resources.

Every item in the knowledge base, regardless of format, is what we call a content that we can return as an answer when users ask questions or search for something on your website.

The knowledge base can contain information in any format you choose in order to provide the best answers for your audience. For example, if your users are looking for products, we can simply provide a link to the products section of your website. On the other hand, if they are seeking solutions to a specific issue we can show them articles, videos, and/or FAQs on how to troubleshoot and solve their problems.
Was this answer helpful?
Thanks!
Meaning-text theory, originally developed by Aleksandr Žolkovskij and Igor Mel’čuk, proposes that the different types of relationships between lexical units (all the elements of a language's vocabulary - words, parts of words, and phrases), can be formally represented as lexical functions.
 
Lexical functions can be used to simplify complex structures and assign semantic weight to the various parts of an expression. This allows linguists (and in turn, our linguistic technology) to relate similar meanings independent of the exact phrasing or word order.
 
Our linguists have adapted the ideas put forward in the theory to create an advanced semantic search, by using lexical functions to describe and simplify natural language expressions with the goal of understanding the meaning of user questions regardless of differences in wording.
Was this answer helpful?
Thanks!
Linguistically speaking, a word is the smallest meaningful independent unit in a language. A word can be used alone and still have meaning, unlike a morpheme (like 'un-' in unlike, which adds meaning but can't be used alone). Words act as building blocks to more complex meaningful statements.
Was this answer helpful?
Thanks!
Disambiguation is the process of resolving and removing ambiguity. In linguistics, ambiguity occurs when a word can have more than one meaning and disambiguation is how the best intended meaning is selected. Every human language has ambiguity: there are words that sound the same (homophones), as in 'steel' vs. 'steal'; words with the same spelling but multiple unrelated meanings (homographs), like 'bow' which can mean a type of knot or a respectful gesture; and words with the same spelling and related but still distinct meanings (polysemes): 'rent' can mean to rent to someone and to rent from someone.
 
Being able to disambiguate between the various possible meanings of a word is essential to understanding natural language, but difficult for computers to do. Inbenta's software uses a variety of tools to disambiguate user questions, allowing us to find accurate search results based on the user's intended meaning.
 
Our linguistic team has created a blog all about the disambiguation process.
Was this answer helpful?
Thanks!
Linguistics is the name for the formal, scientific study of language. The goal of linguistics is to describe how languages actually work; Inbenta's linguists apply scientific knowledge about languages to solve practical problems, like natural language processing.
 
There are many different aspects of language to study, including: phonetics, the study of speech sounds; semantics, the study of meaning; syntax, the study of grammar; morphology, the study of word structure; and more.
 
Computational linguistics is a field in which linguists work with computer scientists to attempt to describe language from a computational perspective. The expert computational linguists at Inbenta come from a variety of linguistic backgrounds, enabling our team to create and maintain dictionaries in many languages.
Was this answer helpful?
Thanks!

Alan Turing proposed in his 1950 article "Computing Machinery and Intelligence" that if a machine could successfully make a knowledgeable observer believe that it is human, it should be considered to be intelligent.

This test satisfies most people for the purpose of determining machine intelligence, but philosophical views of intelligence may differ.

Typically for the Turing test, an observer interacts with a computer through a text interface, typing back and forth so that the machine doesn't have to imitate the sound or appearance of a real person. If the observer believes they have been interacting with another person, rather than a computer program, the computer has "passed" the test.

Was this answer helpful?
Thanks!

Backstage is a web application that our customers can use to track the progress of a project, maintain and edit a knowledge base, process support tickets, and more. It centralizes all the different elements of a support center, so agents can easily manage content, look up answers, and respond to questions with one application.

Contact us for a live demo of how Backstage handles more than 300,000 support tickets and can give you an industry leading self-service rate of more than 90%.

Was this answer helpful?
Thanks!
A web crawler, or spider, is a computer program that browses the World Wide Web in a methodical, automated manner or in an orderly fashion. Other terms for web crawlers are ants, automatic indexers, bots, and web robots.

Inbenta provides its own spider, called NePHPila, that can crawl and index virtually any website, XML stream, or document base to build a searchable knowledge base automatically.
Was this answer helpful?
Thanks!
A tag cloud (also known as a word cloud or weighted list in visual design) is a way to represent the content of websites visually, by making a 'cloud' where common words or tags have a particular font color or size. They give an immediate idea of what topics or words are most frequently used. Even though Inbenta's unique semantic clustering technology allows to cluster groups of similar user search queries, it's not a tool designed to implement tag clouds.
Was this answer helpful?
Thanks!

Inbenta Search is built on a foundation of over 10+ years of R&D invested in natural language processing, which understands what your customers are typing. It then reads between the lines to respond with the correct result over 90% of the time.

Contact us for a live demo of how Inbenta Search can help you or watch our video.

Was this answer helpful?
Thanks!

The bigger, more complex and disparate the data set, the stronger machine learning becomes in getting high quality insights and generating value for your business.

Machine learning gets smarter and digs deeper as the amount  of information it has access to grows, no matter how many variables, changing parameters and platforms. Machine learning sees the patterns in data, so the more data you put into the system the clearer the patterns become.

Was this answer helpful?
Thanks!

Inbenta's patented natural language processing technology will ensure your help site will be able to recognise the customer's natural language and, therefore, understand the meaning behind the words.

The result is an enriched customer experience, reduced support costs and increased revenue.

Check out our five steps to a successful self-service strategy using NLP.

Would you like to see a live demo of how Inbenta's +90% self service rate can help you? Get in touch!

Was this answer helpful?
Thanks!

Natural language processing ensures computers can understand natural language. 

Inbenta has natural language processing, or NLP, at its core. Theoretical linguistic frameworks like the meaning–text theory (MTT) — used for constructing of models of natural language — allow computers, and thus your search technology to process natural language by understanding the meaning behind the words.

Check out our white paper on how NLP works.

Contact us for a demonstration of how our patented technology works!

Was this answer helpful?
Thanks!

Inbenta Hyperchat gives your support agents highly responsive and intuitive tools for helping online customers faster and smarter. For customers it means anytime they’re stuck, a chat window launches, instantly connecting them to that live agent, who have AI-fueled NLP tools to provide answers right away.

Give live support agents more flexibility

Integrate support ticketing with Hyperchat and your support center staff will have greater flexibility to answer customer questions faster and more accurately. Not only does this cut down on the need for email responses, which are often slower, it’ll result in better outcomes, which will enhance the overall customer experience.

Comes equipped with Backstage Analytics

Now your agents can quickly access your website knowledge base, plus live chat is integrated with this powerful analytics management portal and ticketing system, which creates a simple agent experience that allows them to convert conversations to emails. No more need to swap tools, applications or screens, mid-sentence.

Was this answer helpful?
Thanks!

Most e-commerce catalogs are organized using SKUs, or Stock Keeping Units, because they provide unambiguous, unique identifiers for every product. But when customers are in the store ready to buy, they're going to search for products using natural language- meaning their queries will be incomplete sentences, ambiguous phrases, and full of irregular grammar and spellings.

Inbenta Search for E-Commerce tools apply natural language processing principles to ecommerce searching, making it easy and natural for customers to find products. The benefits are enormous; customers have a frustration-free shopping experience, and conversion and click-to-cart rates improve.

Contact us to arrange a live demo.

Was this answer helpful?
Thanks!
Although we don't develop automatic speech recognition technologies, we have partnered with leaders in the industry to be able to offer the best combination of speech-to-text and natural language search. We can provide solutions particularized to specific platforms (mobile vs. web, etc.) and languages - just contact us for more details!
Was this answer helpful?
Thanks!

With Inbenta Surveys, customers can use our analytics and ticketing tool, Backstage, to create surveys, publish them, and see the results. The system is very easy to use and also very flexible. A survey can be linked to a virtual assistant session or a live agent conversation.

One of the most important aspects of customer support and service is to know your customer and to listen to their opinion. One of the best ways to do this is to ask them directly. Using the survey tool, your customers will share their opinion on any topic you would like to know.
 

Get in touch for a free demo of the Inbenta Survey.

Was this answer helpful?
Thanks!

Inbenta can target long-tail searches and generate SEO landing pages to drive traffic to your site.

The long tail represents all the requests that web users type in the search engine that are not very common.

According to Google Trends, the request « How to ensure a dobermann » is a "long tail request" because less than 10 web users are typing this request every month.

Conversely, the very popular request «car insurance» is a "top tail request". More than 100,000 web users are typing this request every month.

To better understand the long tail phenomenon, take a look at this graph:



From this graph, we can see that the vast majority (80% - the green line) of requests made by users are "long tail requests", that is to say, requests that individually are not very often typed into search engines. It shows that everyone has their own way of expressing themselves and formulating their requests on search engines.


In the end, few people are repeating the exact same request. Only 20% of web users' requests are "top tail" requests, that is to say, requests that are very popular such as «Shoes».

We can also learn two other tips from this graph:

  • Long tail requests consist of an average of 4-5 keywords.

  • With 4-5 keywords, web users can articulate their needs more accurately than with only 2 keywords (top tail request). The more a web user is able to express their need with accuracy, the more advanced they are in the decision/purchase process. In other words, traffic coming from the long tail is a higher quality traffic than the one coming from the top tail and this ensures a greater ROI.

  • Our long-tail SEO feature allows you to take advantage of these longer, targeted searches by generating landing pages
Was this answer helpful?
Thanks!
Autocomplete suggests answers from your knowledge base, website pages, and even previously answered user questions as your users type. It analyzes their keystrokes to find the right information faster, delivering information to your users as they are asking questions. The easier and more intuitive it is for your users to find what they want, the more positive their overall experience with your company is.
Was this answer helpful?
Thanks!
Traditional cross selling boosts sales by suggesting additional products to a customer based on their previous purchases, but at Inbenta we've turned cross selling into a conversion rate optimization tool that encourages users searching your knowledge base to learn more about your related products and services, and potentially make a purchase.
 
Fully integrated into our knowledge management system, the cross selling module allows you to create special CRO pages that are displayed based on customer searches. Because they're powered by our natural language search technology, the cross selling suggestions are always accurately targeted to that specific customer's interests. Every time they search for information, the system can suggest related products and services intelligently, maximizing sales opportunities.
 
Contact us for a live demo of how cross selling can increase product sales.
Was this answer helpful?
Thanks!

Inbenta's chatbots can be help you with transactions. We can provide either step-by-step guidance for completing a purchase or even handle the purchase directly, alleviating the visitor of any lingering questions or concerns that might lead to an incomplete purchase.

Nearly 70% of online shoppers abandon their shopping cart before completing a purchase, resulting in nearly $4 trillion-worth of merchandise left waiting in virtual carts to be purchased.

Contact us to find out how our chatbot will give you higher conversions, fewer abandonments and more repeat customers.

Was this answer helpful?
Thanks!

With Inbenta Self-Service Solutions, you can provide all the information your users need, when and where they need it.

Our applications make customer service an engaging experience and streamline customer support processes for everyone - agents and customers alike.

Wherever your users are and whether they access your site from a PC, phone, tablet or any other device, we deliver the same relevant information and quality answers that they're expecting.

If you want to see exactly how our features can help you improve customer satisfaction, reduce customer support and increase revenue then get in touch for a live demo!

Was this answer helpful?
Thanks!
Inbenta comes with a powerful content management system that will allow you to create FAQs and edit your knowledge base in a detailed yet easy-to-understand way. Please fill in this form to arrange a meeting with Inbenta so we can tell you more!!
Was this answer helpful?
Thanks!
A knowledge base can be as big (or as small) as you want, but depending on your needs, we may be able to offer some recommendations for an optimal knowledge base, based on our experience with similar projects.
Was this answer helpful?
Thanks!
Here at Inbenta, we are experts in the art of the FAQ. We've written a guide to writing FAQs that are not only more helpful to your users but can even encourage your users to make a purchase. Please fill in this form and we'll send you our FAQ Editorial Guide.
Was this answer helpful?
Thanks!

We can create a knowledge base for you from your existing content; just send us a spreadsheet or XML file that contains your current contents. We can also index contents directly from your website and upload them into our system.

After the initial upload, you will be able to monitor and update everything using a program called Backstage, our knowledge management system.

If you are using some other platform to manage your contents, we can automatically index your website as many times per day as you want so your contents are always up to date.

Contact us for a live demo to see exactly how this works.

 
Was this answer helpful?
Thanks!

Inbenta will improve your knowledge base through semantic clustering.

Semantic clustering helps your company discover gaps in your content to enrich your customer’s experience.


Inbenta’s Semantic Clustering groups semantically equivalent search queries — words, phrases and sentences — into clusters based on meaning. The higher the number of questions, words and phrases with a similar meaning, the greater the cluster.

This gap analysis function allows our customers to discover where they need to add additional answers and information to make sure their customers can find the information they want quickly and easily, without having to raise a support ticket.

Contact us for a live demo to see exactly how it works.

Was this answer helpful?
Thanks!
Spelling correction is important both for content creators and for search accuracy.
 
Typos in knowledge base articles can confuse users or even give a poor impression of your brand, and misspelled words in search queries make it harder to detect meaning.
 
Our spell-checking tools can identify mistakes as editors create new content, and periodically provide a list of typos in existing FAQs so they can be corrected.
 
For search applications, our spelling corrector analyzes unknown words in user questions and applies corrections, allowing it to complete the search with the intended word.
Was this answer helpful?
Thanks!
Inbenta is an SaaS (Software as a Service) provider, meaning we host content for you. Although we store your content and make it easy to deliver it to your customers, all the content and data that we host for your project still belongs to you.
Was this answer helpful?
Thanks!

Inbenta Knowledge Management makes it easy to have all the information you need in one place. Control your information and make sure that your customers can ask questions and get the right answer.

Data becomes more valuable every day and being able to utilize it to make informed decisions is crucial. Inbenta Knowledge Management puts the data you need at your fingertips because Inbenta uses natural language processing (NLP) to truly understand information tagging, classifying and grouping; content and answers are automatic.
Easy access to structured data makes every part of your organization better informed and more agile.

 

Was this answer helpful?
Thanks!

We understand the need to compartmentalize the information in your knowledge base, so we created a system of profiles that you can use to control which users can view specific content. You can display or hide specific articles based on what part of your website a user comes from and whether or not they are logged in. As well as keeping internal documents separate from public user guides. It's easy to customize access groups to your exact specifications because Inbenta gives you total control over who can view your content.

Contact us to arrange a demo.

Was this answer helpful?
Thanks!

The technology we've developed at Inbenta makes tags and metadata obsolete. Developed by an expert team of computational linguists, our search engine finds results based on the actual meaning of the input it receives, not by matching to exact text strings. Keywords, metadata, and tags are unnecessary because the software is able to process natural language.

Contact us to find out how our patented natural language processing technology can help you.

Was this answer helpful?
Thanks!
The Semantic Web is a set of standards that allows for data interoperability on the internet using XML formats, with the goal being to create a "web of data". Searching the Semantic Web requires data searches, rather than web page searches. Commercial use of the Semantic Web is very limited, although Inbenta already allows the creation of semantic functions for our clients' websites and thus reduces customer service expenses and increases online sales.
 
Was this answer helpful?
Thanks!
The linguistic matching process between user questions and FAQs can use information from any text field (the title, the answer, the category, etc.). As part of the setup process, our linguistics team will calibrate our tools to prioritize the fields on your site that contain the most meaning for matching purposes.
 
With our powerful semantic search technology, it's not necessary to tag content or use metadata; our software can extract meaningful information from your contents automatically, making it fast and easy to implement. If the information is there, Inbenta can find it.
 
The most important thing for your customer support center to focus on is creating high-quality, understandable answers that respond to the needs of your users. And no worries, because Inbenta will provide feedback and suggestions to help you better adapt your knowledge base to your customers!
Was this answer helpful?
Thanks!
If our software doesn't recognize a word in a user question, a few things can happen. First, the software will apply spelling correction modules; if the input is corrected to a real word, the search is completed. If the input isn't corrected to an existing word, the software will search your knowledge base for the word. If you've added new terminology like product or brand names, the search will still be able to match to your new words. If the word isn't a correctable spelling mistake and isn't in your knowledge base, our linguistics team will analyze the word and decide whether it should be added to our dictionary.
 
Our dictionary doesn't need to contain every word, just words that carry real semantic weight. Unrecognized words don't always need to be added because if they don't contribute real meaning to the query, we don't need to analyze them.
Was this answer helpful?
Thanks!

Inbenta's Semantic Clustering helps your company discover gaps in your content to enrich your customer’s experience.

When a customer does not receive a satisfactory answer they will either leave the page or fail to click on the content offered. Semantic clustering alerts you to this negative response and to the need to create new material to answer the question. 

This is a huge advantage as it enables businesses to see what their customers want to know and respond with better-targeted content that improves customer satisfaction.  

Additionally, our team analyzes user questions with no answer regularly and compiles reports to inform you of the most common unanswered topics, meaning your knowledge base constantly adjusts to your users' questions. As new issues arise, your support center can always respond quickly and efficiently.

Contact us for a live demo of exactly how our semantic clustering can significantly reduce your unanswered questions and support tickets.
Was this answer helpful?
Thanks!

Semantic clustering helps your company discover gaps in your content to enrich your customer’s experience and reduce the number of unanswered questions.

Inbenta’s Semantic Clustering groups semantically equivalent search queries — words, phrases and sentences — into clusters based on meaning. The higher the number of questions, words and phrases with a similar meaning, the greater the cluster.

This gap analysis function allows our customers to discover where they need to add additional answers and information to make sure their customers can find what they want quickly and easily, without having to raise a support ticket.

Contact us for a live demo of how semantic clustering can increase customer satisfaction.

 
 
Was this answer helpful?
Thanks!
Semantic search prioritizes the meaning of an input rather than searching for exact text strings (like keyword search does), meaning it matches relevant answers to user questions even if the phrases used in questions and answers are totally different.

Developed by our expert linguistic team and based on meaning-text theory, Inbenta's Semantic Search technology is capable of finding incredibly accurate results no matter how users word their queries, without needing any extra work from your content creators and editors. There are no tags or metadata involved; our extensive dictionary and catalog of lexical functions allows our software to determine the intended meaning of any query. You simply create content, and our software does the rest!
 
 
Was this answer helpful?
Thanks!
The biggest challenge that stands in the way of computers understanding natural language is ambiguity. Natural languages are filled with ambiguity; words and phrases can have identical form and/or the same grammatical structure, but very different meanings. In the same vein, one concept or idea can be expressed in a variety of ways by paraphrasing.

Take the word charged, for example: in the sentences "He charged the phone", "He charged the customer", and "He charged the enemy", it has totally different meanings. A human reading or hearing these sentences will know immediately that "charged" means something like 'replenished the battery' in the first case, 'billed costs to' in the second case, and 'attacked' in the last case. The meaning of "charged" is very different in each context, but a computer analyzing these sentences will find similar grammatical contexts and will struggle to disambiguate between the different possible interpretations of charged.

Another big problem in natural language is paraphrasing; there are almost always multiple ways to express the same concept. A person who has purchased a shirt that doesn't fit might say "I'm too tall for this top", "it fits tighter than I expected", "the blouse doesn't fit me", "the shirt's the wrong size", "this tee runs small", or any other of a multitude of possible phrases. All of these are expressing the same concern, that the shirt is not the correct or desired size, but the wording used is totally different.

Humans are very good at determining the correct meaning in ambiguous cases using contextual knowledge about the world and their specific language and we're also very good at recognizing different phrases that have the same meaning, but it's very difficult for computers to handle ambiguity and paraphrasing.

Our technology uses extensive dictionaries that describe words as well as all the different possible relationships between them, called lexical functions, to resolve ambiguous words and phrases so that the software can 'understand' natural language. Semantic search by Inbenta identifies the meaning of specific words and phrases, allowing it to match to results by concept, rather than keywords.
Was this answer helpful?
Thanks!