Logana Product Strategy

The members of Logana Information Research Center have decided to create the Logana Product Family on the basis of the Logana Information Research Technology developed by them, through which they want to ensure the spread of this new technology in the world.

A demonstration program of this technology can be tested on the Internet; www.logana.com.

Note that this demonstration program has two functional pages (called Analog1 and Analog2) and both have an English and a Hungarian version. On the Analog1 page, we can perform a similarity search by distorting one of the names selected from the internal name set of the presentation program. The Analog2 page allows you to examine your own text (up to 15,000 words), where the text can be copied for example, using Ctrl-C after text selection and thereafter Ctrl-V. Here, in addition to similarity text search, you can also perform automatic keyword collection. (It is a good idea to read the Help before use!)

What is Logana Information Research Technology?

Logana Information Research Technology is a set of word and data processing methods based on a completely new interpretation of similarity that is uniquely efficient in the world and enables text search, data collection, and content analysis in a language-independent manner.

Other applications include automated keyword extraction, computer virus search, and can help with text document analysis, comparing texts, hunt for plagiarism, phoneme-based speech recognition, searching for mutant gene sequences in the DNA chain, etc.

Below we present the main areas of use and potential applications of our technology. By implementing these, the search, interpretation, analysis, classification, collection, and processing of textual (and generally sequentially structured) data in our world will change significantly.

1.) Primary Applications

These include applications, the use of which is well illustrated in our demonstration program (www.logana.com) and can already be operated with a relatively small infrastructure background.

1.a.) Logana Content Analysis System (Logana CAS)

(Support of document processing)

State and security institutions, large companies and libraries often need to examine, interpret and process large amounts of text documents.

The prominent element of the Logana Product Family is the Logana Content Analysis System, which enables Automatic Comparison, Automatic Content Analysis (Verification) of Texts, Text Documents, the Automatic Keyword Extraction, the Automatic Topic Classification, etc. So the "what is it about?", or the "how similar is it?" questions about any document (or documents) can be easily answered.

The basic element of this system is the Similarity Text Search, which can be used to find those incorrectly searched and/or incorrectly stored or multi-shaped words also, for which the traditional web search engines (Google, Bing, Ask, Yahoo, etc.) are completely unsuitable. (See eg Satöbrian [51%] → Chateaubriand, Feiszekker [60%] → Weizsäcker on the English Analog1 site, or Solnoki [60%] → Cholnoky, Széscényi [80%] → Széchenyi on the Hungarian Analog1 site.) This method is particularly useful in case of long and easily misspelled words.

Finally, it should be noted that this method of analysis can be used to investigate plagiarism, to prepare for a general psychological examination, and to identify sources of information suspected of crime or terrorism.

1.b.) Logana Phonetic Teaching System (Logana PTS)

(Support for language teaching)

This enables language learners to search a given language’s dictionary and textbooks for words written according to their pronunciation, thus simultaneously ensuring that they learn the correct spelling of the word they are looking for, as well as its meaning and application in the given language.

1.c.) Logana Similarity Library System (Logana SLS)

(Support for Library Usage)

The Logana Similarity Library System provides the following

– for library readers, the Similarity Text Search presented above in the documents of a library (or even a national library network!),

– for librarians the Automatic Keyword Extraction from documents in different languages,

– for librarians the creation of a Reference Glossary for automatic keyword extraction in any language and specialization.

2.) Perspective applications

These applications require a high-performance infrastructure (server park).

2.a.) Logana Similarity Textbook Searcher (Logana STS)

The Logana Similarity Textbook Searcher, available nationwide, enables elementary school students to use similarity search in textbooks with possibly misspelled search words, thereby ensuring the student’s apprehension of both the correct spelling of the search word and the related educational material. It is advised to start the introduction of this system with the Literature, History, and Geography textbooks.

2.b.) Logana Similarity Web-Searcher (Logana SWS)

This is the most spectacular application, because this is able to find those (incorrectly searched and/or stored or multi-shaped) words, that the traditional web search engines (Google, Bing, Ask, Yahoo, etc.) can't find (see Examples 1.a.).

As this requires a very powerful server background, it is recommended to limit initial researches to a particular field. (For example “Cultural programs in the country”, “American writers”, “Nobel laureates in the world”, “Food recipes”, etc.)

2.c.) Logana Similarity Semantic Searcher (Logana SSS)

It allows you to perform a Similarity Text Search not only for the word you are looking for, but also for all its Cognate or Opposite Meanings (Synonym or Antonym), thus ensuring Search by Content in the language(s) selected by the user. An example for multi-meaning words or synonyms: Case (→an example or instance of something), (→a legal matter that will be decided in a court), (→a container or cover for keeping or protecting something). Simply put: if we’re looking for Dogs, we should find Hounds as well.

This is how it works:

– as a first step, it offers words that are "sufficiently similar" to the word the user is looking for, but have different meanings, and give each of them its meaning, so that the user can choose the right one and then

– in the second step, it designates (in the document or database under investigation) all “sufficiently similar” shaped hits to all Synonyms (to synonymous versions), or Antonyms (versions with an opposite meaning) of the word selected in step one, depending on which option the user chooses (synonym or antonym working mode).

Finally, it should be noted that although Logana Similarity Text Search is language-independent, semantic search is required for the selected language(s)

– an interpretive dictionary (for the first step) to allow the user to select the word that has the appropriate meaning, and

– a synonym dictionary or an antonym dictionary (for step two).

2.d.) Logana Similarity Semantic Web-Searcher (Logana SSWS)

By combining the above two systems (Logana SWS and Logana SSS), you can create the world's first Content Search on the Internet.

3.) More Applications

The creation of these applications requires intensive cooperation with representatives of other scientific disciplines.

3.a.) Logana Gene Search System (Logana GSS)

The Logana Gene Search System allows automatic search for the mutant gene sequences in a DNA chain.

Thus, the geneticist only needs to examine (interpret) those DNA chain portions where the result of the Framework Relative Comparison reaches a predetermined Limit-Similarity value. Naturally, this Limit-Similarity value is given by the geneticist (as a percentage), and he or she can search multiple times for different Limit-Similarity values.

This is a rather resource intensive process (a DNA chain consists of billions of triplets!), but on the one hand a computer of this computational power doesn’t come at an astronomical price, and because this process is automated, therefore it is much more efficient and reliable than the conventional chemical method.

3.b.) Logana Speech Recognition System (Logana SRS)

The Logana Speech Recognition System also significantly facilitates the understanding of faulty speech (stutter, lisp, etc.). A prerequisite for its use is the conversion of the speech to be examined into a phoneme-sequence. Phonemes are the “letters” of the spoken language, and obviously they can vary not only by language, but also by dialect, and people with faulty speech may develop other distortions.

The procedure consists of the following two steps:

– In the first step, an initial, not too long part of the speech is examined by similarity processing after a speech-phoneme conversion to determine the language (and dialect) of the speech.

– In the second step, even the faulty speech phoneme sequences can be successfully interpreted by the similarity processing applied to the speech to be examined after the speech-phoneme conversion belonging to the language (language procedure) selected in the first step.

3.c.) Logana Similarity Prefiltering Hardware (Logana SPH)

The search for similarity of sequential structures (e.g. text search in a document, table or database) can be significantly accelerated by (FPGA or GPGPU) hardware pre-filtering. As a result of this, the more resource-intensive part of the similarity process only needs to be performed for a significantly smaller set of sequences.

Other potential applications include: melody comparison, search for similar melodies, plagiarism checking of texts, similarity-based text-correction, morphological dictionary, automatic text interpretation, general psychological text analysis, etc.