FAQ - User

[2.1] Which are the features that my browser should have in order to enjoy all the interface qualities?

Although you can access to the system with any browser with frames support, the system will work better if your browser supports Javascript and CSS.

[2.2] Would I add my own information by myself?

Yes you would, in the user's manual you can get the necessary information for adding authors publications and references.

[2.3] Would I get an interface in my native language?

No and there are no plans so that this happens, mainly because almost all the scientific information appears in English language therefore we think that it is not necessary to offer the interface in other languages.

[2.4] Why ScienSe does not identify all the information of the articles correctly?

There exits many reason by which ScienSe is not able of identify all the information of an article correctly, the most commons are:

1. The format of article is not common. ScienSe tries to identify all the information of an article using probabilities, so, it is designing for identify the information in the most common format, when this is not true, ScienSe can not know which is the sort of the fields, so it gets confused.
   
   Altough this is not a problem that affects the most quantity of the articles we hope to improve notably by using a neural network (see Future Projects for more information about it).
   
   
2. The programs that helps to convert the file, from any of the supported formats, to text can no convert the file of the article correctly. ScienSe can support files with different formats as pdf, ps, rtf and text, but in order to work with a file, ScienSe requieres it to be as a plain text file, so it uses some format converters that get those file as plain text files. Unfortunatly, there are some programs that save the files with some no standar characters or format, so the converters can not get all the information from the files with that characteristic. Thus ScienSe receives a damaged file different from the original.

[2.5] I have install a new version of the converter filter, why does ScienSe seem to have changed?

ScienSe can support different formats but in order to execute the algorithm for identify the information it uses some converters format (as pdftotext) so as to get all the files as plain text files.

When you install a new version or a different converter, the plain text files might changed also, since the program which converts them have changed, so, the file given to ScienSe could be diferent, and this could change the results.

[2.6] Which is the "ideal" format, that ScienSe requieres for getting the general information of an article?

As there are a lot of different ways for writing an article, there is no way to be sure about the sort of the information, so in order to improve the performance of ScienSe is desirable that the files of the articles fulfill certain formats. The desirable format is:

1. The first line of the file should be the title of article, and it should finish with an endline (enter).
   
   
2. The authors should be separated by "," (commas) except for the last two authors which should be separated by either the word "and" or the character "&". For example:
   
   Raggi Francisco, Urrutia Jorge, Miramontes P., Bracho J. and Viso Elisa for meaning: 5 authors.
   
   
3. If the article contains the month of publication the entire word in english should be writen instead of any abreviation. For example: January, and no jan, or enero, or something like that.
   
   
4. The emails, if there is some, should be set in the same order that the authors where writen, and if there are some authores with no email address they should be set at the end. The email addresses can be separated by ',' (commas) or only with ' ' (spaces). For instance:
   
   Raggi Francisco, Urrutia Jorge, Miramontes P., Bracho J. and Viso Elisa fraggi@matem.unam.mx, urrutia@matem.unam.mx pvm@fciencias.unam.mx
   
   for meaning:
   
   Raggi Francisco fraggi@matem.unam.mx,
   Urrutia Jorge urrutia@matem.unam.mx,
   Miramontes P. pvm@fciencias.unam.mx,
   Bracho J.
   Viso Elisa
   
   By simplicity there should be at most one email address by author.
   
   
5. The personal web page should be in the same order that the authors were given, and can be separater either by and ',' or a ' ' character. For instance:
   
   Raggi Francisco, Urrutia Jorge, Miramontes P., Bracho J. and Viso Elisa fraggi@matem.unam.mx, urrutia@matem.unam.mx pvm@fciencias.unam.mx http://www.matem.unam.mx/investigadores/fraggi/ http://www.matem.unam.mx/~urrutia/ http://www.mathmoo.unam.mx/biomat/pedro/pedro.html, http://www.matem.unam.mx/investigadores/roli/,
   
   
6. The abstract should begin with the word "abstract" and should finish with an end line character (enter). For instance:
   
   "Abstract. This is an example."
   
   or
   
   "abstract
   This is an example"
   
   
7. The phones should be a string of numbers with at least 5 numbers, separated by the character '-'. And they should be given in the same order that the authors. For instance:
   
   Raggi Francisco, Urrutia Jorge, Miramontes P., Bracho J. and Viso Elisa 50-00-00-00 51-11-11-11 01-23-45-56-78-9 98-76-54
   
   for meaning:
   
   Raggi Francisco 50-00-00-00,
   Urrutia Jorge 51-11-11-11,
   Miramontes P. 01-23-45-56-78-9,
   Bracho J. 98-76-54,
   Viso Elisa ,
   
   
8. The name of the journal or magazine should be given using any of the following words, and should finish with an end line (enter):
   
   
   1. "journal". Example: Linux Journal
   2. "magazine". Example: Science Magazine
   3. "communications". Example: ACM Comunications
   4. "scientific". Example: Scientific American
   5. "proceeding". Example: Proceedings of the National Academy of Sciences
   6. "book". Example: Algebra's book
   7. "review". Example: ACM Operating Systems Review
   8. "nature". Example: nature.com
   9. "science". Example: Science Online
   
   
9. References section should the last section of the acticle, and should begin with the word "references" (it does not matter if the letter are upper or lower).
   
   
10. 10.- Each reference should begin with one of the following ways:
    
    
    1. [anything] anything
    2. (anything) anything
    3. {anything} anything
    4. anything.- anything
    Where anything means any character, number or letter different from the ones that contains it. NOTE: the first reference shows the way in which the rest of the references are going to be waited, so if the first begin with [1] the next references should be given with [ and ]. For example:
    
    [1] Marvin Minsky, "The Society of Mind", Simon and Shuter, 1985, 300p.
    
    [2] Marvin L. Minsky and Seymour A. Papert, Perceptrons: Introduction to Computational Geometry, (1988), MIT Press
    
    or
    
    (1) Marvin Minsky, "The Society of Mind", Simon and Shuter, 1985, 300p.
    
    (MM) Marvin L. Minsky and Seymour A. Papert, Perceptrons: Introduction to Computational Geometry, (1988), MIT Press, 34-40.
    
    or
    
    {1} Marvin Minsky, "The Society of Mind", Simon and Shuter, 1985, 300p.
    
    {a} Marvin L. Minsky and Seymour A. Papert,
    
    Perceptrons: Introduction to Computational Geometry, (1988), MIT Press, 34-40.
    
    or
    
    SM.- Marvin Minsky, "The Society of Mind", Simon and Shuter, 1985, 300p.
    
    1.- Marvin L. Minsky and Seymour A. Papert,
    
    Perceptrons: Introduction to Computational Geometry, (1988), MIT Press, 34-40
    
    

[2.7].- Which is the desirable format for give the references of an article according with ScienSe?

Each cite should be given with the following format:

1. The pages of the journal where the article is, should be given either by two numbers separated by a '-' or by a number finished with "p."
   
   Example:
   
   [SM] Marvin Minsky, "The Society of Mind", Simon and Shuter, 1985, 300p.
   [P] Marvin L. Minsky and Seymour A. Papert, Perceptrons: Introduction to Computational Geometry, (1988), MIT Press, 34-40
   
   
2. The year of publication should be older to 1900 and younger to the present year (of the computer where ScienSe is running). The year can be given as the number alone or as (number) with the '(' and ')' characters.
   
   Example:
   
   Marvin Minsky, "The Society of Mind", Simon and Shuter, 1985 for getting:
   
   year=1985
   
   or
   
   Marvin L. Minsky and Seymour A. Papert, Perceptrons: Introduction to Computational Geometry, (1988), MIT Press
   
   for getting:
   
   year=1988
   
   
3. The journal should be the last string of the reference after erasing the year and the pages. The journal is consider as all the string before the last ',' (comma) character, so the journal should no contain any ',' character within the string.
   Example:
   
   Marvin L. Minsky and Seymour A. Papert,
   Perceptrons: Introduction to Computational Geometry, MIT Press
   
   for getting:
   
   journal = "MIT Press"
   
   
4. The title should be the last string of the reference after erasing the year, pages and journal. The title is consider as all the string before the last ',' (comma) character, so the title should no contain any ',' character within the string.
   Example:
   
   Marvin L. Minsky and Seymour A. Papert,
   Perceptrons: Introduction to Computational Geometry
   
   for getting:
   
   title=Perceptrons: Introduction to Computational Geometry
   
   
5. The authors are consider as contained in the string of the reference after delete from it the year, pages, journal and title. Each author is separed by the ',' character from the others and can contain the word "and" or the character '&' separating the last two authors. NOTE: as it is supose that each author is separated from the others by a ',' the any name of author shoul contain the ',' character. Example:
   
   Marvin L. Minsky and Seymour A. Papert
   
   or
   
   Raggi Francisco, Urrutia Jorge, Miramontes P., Bracho J. and Viso Elisa
   
   or
   
   Raggi Francisco, Urrutia Jorge, Miramontes P., Bracho J., Viso Elisa
   
   

[2.8] Is the ideal format the unique format that ScienSe is able to work with?

No. ScienSe work reasonably good (it gets correctly up to the 50% of information in an article) with almost any format, but the error's probabilities increase when the articles do no have an usual format.