Tayait.com Goes Live
Offering features never seen before in the realm of Arabic Search
The Arabization of the internet has made huge leaps and bounds forward over the last few years, with major content providers amassing massive amounts of traffic. Yet one major hurdle has remained – a quality service providing users the ability to easily and accurately search the Arabic content available on the internet today. Though recently we have seen a few sites moving into the realm of Arabic search, none have attempted to go back to the basics and deal with the complexities of Arabic as a language in order to improve their search results. It is this complexity that has left major players like Google, even after providing for searches in languages as diverse as Yiddish, Swahili, Twi, and even Pig Latin, lagging behind when it comes to Arabic search.
This is where tayait.com has diverged from the rest of the pack. Spending over a year in collaboration with Britannica developing an Arabic Natural Language Processing (NLP) module that offers users a level of Arabic processing that has never been seen before, tayait.com offers an Arabic search function that now allows computers to “understand“ the language sufficiently to comprehend the ways in which the language may be used.
NLP is a science that combines Artificial Intelligence, language properties, and statistics and it has been used to address the main two problems concerning the Arabic language: high inflection and the absence of short vows.
1) High inflection: Arabic is a highly inflected language, which means that from a single root a large number of different forms can be generated.
2) The absence of short vowels within written Arabic creates high ambiguity, where the word being referred to must be inferred by the reader from its context. This creates a level of complexity that we as humans have no problem in understanding, but for a computer it becomes incredibly difficult to process.
taya it is introducing tama (taya it Arabic Morphological Analyzer) as native plug-ins that are easily integrated into leading enterprise and web search engines including: Oracle text, Sun Solaris, Convera RetrievalWare, Microsoft SQL Server, SharePoint, and FAST. tama helps dramatically search in Arabic through the NLP technology.
taya it.com offers users the ability to search in the following language modes:
• Morphological Search: results include all inflections of the keyword.
• Expanded Search: results include all derivatives of the keyword.
• Synonyms (Thesaurus): returns all synonyms and dialectical variants.
• Cross Language: the ability to input English words to search Arabic documents
• Phrase: the ability to link multiple keywords more loosely or tightly.
At the same time it allows users to “refine” their searches based on automatically generated related keywords, allowing users to drill down as deep into the Arabic internet as they like.
taya it has already realized that without quality websites writing great content their search engine is useless. According to Wafik Shamma, taya it chairman, “there is a huge opportunity on the internet - opportunity for websites employing hundreds of individuals, to boys and girls sitting in their rooms with dreams of being heard; we plan to connect these websites to end users and effectively connect them together as well. We want to create a publishers’ network, making it easier for websites to obtain the resources and advertising revenue they need to be successful.”
With an extensive database of Arabic websites, coupled with superior NLP software, and offering context-based advertising options, this small startup is definitely something to keep your eye on over the next couple of months, as it looks like it has the potential to really shake up the online industry in the Middle East.
Visit their website at http://www.tayait.com to check it out!
- Taya IT Enterprise Solutions Seminar in Dubai
- Taya IT Announces Second Beta Version of Yajeel.com
- Taya IT announce the beta release of Yajeel.com
- maktoob group showcases araby the world's first arabic search engine at gitex 2007
- Madar Research Joins Multinational Effort to Develop Powerful Arab Search Engine