Potential Applications
Understand the potential of the Kudos Semantic Platform & API through an exploration of example Semantic Applications & potential development projects.
Example Applications & Development Opportunities with the Kudos Semantic Platform API
Explore some Example Applications
The Semantic API has the capability to power a range of innovative applications. This page explores possible ideas and concepts for applications. Each concept is provisional in nature and simply gives an indication of a suggested direction.
‘Pages Like This’ or ‘Related Pages’ Widget
Concept: Create a web interface that allows users to create a HTML code snippet that they can add to any site (or CMS), which will automatically display related web pages from a specific site that are similar in meaning to the one currently being displayed in the browser.
This would allow site owners to add a ‘Related Pages’ feature to every page on their site, that would dynamically and automatically show other pages on their site that had a close relationship. But remember it could also be added to a 3rd party site, let’s say site X that shows related pages from site Y. So it could be used to show relationships between sites as well as within a single site.
Remember a web page can be about anything: a news article, an event, a product, a person etc, and therefore depending on the web page currently being viewed, this widget would show related articles, events, products, or even people from the same (or different site). Such a widget would represent an extremely versatile and powerful product.
The user would simply provide the URL of the website they want to show related pages from and the web interface would provide the HTML code (snippet) to be placed on their site or a 3rd party site.
Technical: The HTML snippet contains javascript that would call the Semantic api via the URI as follows: “api/content/like/content/” + hash(window.location)
‘Books Like This’ or ‘Related Books’ Widget
Concept: Create a web interface that allows users to create HTML code (a snippet) that they can add to any site (or CMS), which will automatically display books that are similar in meaning to the web page currently being displayed in the browser. This could link to any 3rd party book website or information source, such as one of the big book websites, or to a vendor such as Amazon, that allows the website owner to gain commission from sales.
Remember a web page can be about anything: a news article, an event, a product, or person etc, and therefore depending on the web page currently being viewed, this widget would show books that related to that article, event, product, or person. This would be very interesting.
The user would simply provide the URL of the website they wanted to show related books from (e.g. Amazon, book website) and the web interface would provide the HTML code (snippet) to be placed in any other site’s template or web page.
Technical: The HTML snippet contains javascript that would call the Semantic api via the URI as follows: “api/content/category/book/like/content/” + hash(window.location)
‘Products Like This’ or ‘Related Products’ Widget
Concept: Create a web interface that allows users to create HTML code (a snippet) that they can add to any site (or CMS), which will automatically display products that are similar in meaning to the web page currently being displayed in the browser. Remember the current web page could be a product itself, but it might be an article, event, or about a person. If you just wanted to use it for products, you could just add it a site’s product template and it would only be used for products.
The product repository could come from the same site, another site, or could link to a vendor such as Amazon, that allows the website owner to gain commission from sales.
The user would simply provide the URL of the product website they wanted to show related products from and the web interface would provide the HTML code (snippet) to be placed in any other site’s template or web page.
Technical: The HTML snippet contains javascript that would call the Semantic api via the URI as follows: “api/content/category/product/like/content/” + hash(window.location)
‘Music Like This’ or ‘Related Music’ Widget
Concept: Create a web interface that allows users to create HTML code (a snippet) that they can add to any site (or CMS), which will automatically display music that is similar in meaning (based off the lyrics) to the web page currently being displayed in the browser. This could link to any 3rd party music website or information source, such as iTunes, that allows the website owner to gain commission from sales.
Remember a web page can be about anything: a news article, an event, a product, or person etc, and therefore depending on the web page currently being viewed, this widget would show music lyrics that related to that article, event, product, or person. This would need to work off a lyrics database to identify songs that were similar and then provide links to a third party music site.
The user would simply provide the URL of the website they wanted to show related music from (e.g. Amazon, iTunes) and the web interface would provide the HTML code (snippet) to be placed in any other site’s template or web page.
Technical: The HTML snippet contains javascript that would call the Semantic api via the URI as follows: “api/content/category/music/like/content/” + hash(window.location)
Find me a book (based on my favourite books or those I’ve already read)
Concept: Create a web interface that allows users to create HTML code (a snippet) that they can add to any site (or CMS), which will automatically display books that are similar in meaning to either the books they have already read or to books they have indicated are favourites. The question here is where to get the list of favourite books from? If this was on a book website, it could be sourced from books they had read or rated. Alternatively with a social plugin, it could examine their favourite books as listed on Facebook and they get related books. Having said that it would most probably be suited to a website that had a list of the books the user had read or had said they liked. This could link to any 3rd party book website or information source, such as one of the big book websites, or to a vendor such as Amazon, that allows the website owner to gain commission from sales.
The user would simply provide the URL of the website they wanted to show related books from (e.g. Amazon, book website) and the web interface would provide the HTML code (snippet) to be placed in any other site’s template or web page.
Technical: The HTML snippet contains javascript that would call the Semantic api via the URI as follows: “api/content/category/book/like/mybooks|myfavourites/” + hash(window.location)
‘News Articles Like this’ or ‘Related News’ Widget
Concept: Provide a pre-configured HTML code snippet that can be added to any site (or CMS), which will automatically display news articles that are similar in meaning to the web page currently being viewed. This could use the Enliten article database, which covers most main stream news sources on the web, to show related articles. This could be a way of informing people of current news on a subject and ensuring that the web page that is currently being viewed, keeps informed of latest news on that subject.
The user would simply be provided with the widget code to be placed in any site’s template or web page.
Technical: The HTML snippet contains javascript that would call the Semantic api via the URI as follows: “api/content/category/articles/like/content/” + hash(window.location)
Brand Analysis
Concept: A company wishes to gauge the general public feeling towards its brand, e.g “apple”.
Approach:
Create a channel for the brand
Attach a twitter search for their brand. All tweets containing their brand name will be output through the feed.
Attach general news sources (RSS) to the channel, filtered by the brand name.
Analyse the channel
Request the channel feed
Compute the relative frequency of semantic tags that have been assigned to each item
Save the results with a time stamp.
Repeat process every day.
Display the analysis
Draw a graph showing the changing pattern of the semantic profile over time.
Brand Comparison
Concept: As per the above brand analysis, but several brands could be compared.
Dating Application
Concept: Create a dating application, widget or plugin that takes 2 sets of data. In one scenario, let’s say the widget sits on a dating website. You ask the widget to compare you to a potential person you want to ask out on a date. Or you ask the widget to find you a potential partner that has similar interests to you. The widget analyses your interests and likes and then potential dates’ interests and likes and suggests people to go out with.
In another scenario, let’s say the dating website allows you to connect your Facebook accounts. Now the widget can use the social data of 2 people to see if they are suited for love. There is potential to add dating or psychological algorithms to see if 2 people are suited based on similarities or not.
Check Similarity / Compatibility of 2 people
Concept: See which people or friends are most like you. Create an app that analyses the social data of you and another person or friend and then tells you how similar you are. This could be used to check for employee compatibility before employing a person.
Experts on This
Concept: Create a web interface that allows users to create a HTML code snippet that they can add to any site (or CMS), which will automatically display experts on the subject matter of the web page currently being displayed in the browser.
The experts could be sourced from social media topic influencers or could be sourced from any experts site. For instance, if used for IT support, the widget could find source experts from an IT expert site and it would display experts relating to the current pages content. This could be done for any topic matter.
Music Discovery Service
Concept: Provide a service that allows music to be searched or browsed by the semantic content of the lyrics. Could be integrated with other music discovery systems already on the market (eg http://www.musicovery.com)
Approach: Purchase a lyrics database (eg http://www.odditysoftware.com/page-datasales1.htm) and import all the songs into concept space.
It can then be queried semantically to find music similar to other music, or music similar to an arbitrary search query: api/content/category/music/like/content/
Connect each music record with a third party music player so that customers can listen to samples of the music they discover.
Connect each music record with a third party vendor so that customers can purchase music they discover.
Possibly rank all the songs by similarity to some concept knoms such as “Emotion”, “Nature”, “Sex”? This might enable us to use these as co-ordinates on a chart (such as that used in Musicovery).
Document Management
Import all your documents into a Semantic Platform database.
Categorise several documents.
Allow the system to dynamically auto-categorise the remaining documents.
Skype Bot
Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.
Legal Precedents
Find legal documents that cover similar issues by placing them all into concept space.
Gmail Fuzzy Tagger
Define a set of “fuzzy” tags. The user wants all emails to automatically assign themselves to the appropriate tag. Each tag is set up as a channel which can be manually configured, OR Each tag learns when you assign an email to it.
Famous Quote Suggester
A plugin for Word, Outlook, Gmail, Firefox, etc that analysis the last few sentences you have just written and suggests a famous quote that might be relevant.
Email Suggester
Every feel like you’ve written the same or very similar email before. This email plugin would suggest similar emails to you as you write one, that way if you written something similar before it can be copied and used again.