Release Notes

Listing of all the techy updates

RELEASE NOTES OF DEEP LEARNING MODULE

Improvements

  • Entity Tagging from NLP with Deep learning for Organization, Job, Degree, and Institutes
  • PDF conversion improvement for column and parallel formats
  • Multiple keyword block heading tagging and splitting (e.g Education, Certification and Trading will be considered as education block)
  • USA address city, state extraction improvement
  • Combine/complete multi-job profile in employment (e.g. Manager/Team Lead/Administrator)
  • Tagging/Extracting institute, degrees with fuzzy logics
  • Improvement in Institute and degree name completion
  • Improvement in tagging/extracting employer w.r.t Location, Job profile and Job period
  • Improvement for address city for UK when the city is present in multiple countries
  • Improvement for address w.r.t UK ZIP code
  • Improvement for Singapore and Hongkong addresses

Bug Fixing

  • PDF header text issue fixing
  • Start and end date fixing in education when education period is a current period (e.g 2018 - Present)
  • Fixing entity issue for apostrophe('s) for Degree, Institute, Sub-institute, and Employer

New Technologies Used

  • Tensor-flow
  • Spacy
  • Word2vec/Glove
  • CRF (Conditional Random Fields)
  • RNN (Recurrent Neural Network)
  • Backward Propagation
  • Gradient Descent
  • LSTM (Long short-term memory)
  • Graph DB
  • Weka NER
  • Maximum Entropy
  • HMM
  • Brat
  • Text Normalization
  • Data Normalization
  • Pattern matching
  • Tagging
  • Fuzzy matching
  • Case analysis
  • Order analysis
  • Delimiter analysis
  • Length analysis
  • Domain analysis
  • Gap analysis
  • Density analysis
  • Semantic analysis
  • Sentence Analyzer
  • NLP
  • Machine Learning
  • Image Analyses (Face Detection)
  • Language Detection
4 August 2020

Latest Updates 

  • Organization Name Extraction from Resumes Improved
  • Chinese Job Period Extraction Improved
  • Employer Location Extraction Improved
  • Employer Profile Mapping Improved
  • Fixed German Executive and Management Summary
  • Improved Employer-Profile Extraction If Defined Without a Date Range Like April 2014, June 2019, etc.

 

16 July 2020

Latest Updates

  • Accuracy of Spanish Resumes Improved.
  • Total Experience in the Year/Month Extraction Improved.
  • Chinese and Spanish Job Period Extraction Improved.
3 July 2020

Latest Updates 

  • LinkedIn Resumes Accuracy Improved
  • Europass Resumes Accuracy Improved
  • Education and Experience Extraction for Spanish and Portuguese Language Resumes Improved.
24 June 2020

Latest Updates 

  • Bug Fixing Done for the End Dates if given as “May June 2019”.
  • Job Period Extraction Bug Fixed.
  • Education Extraction Improved for
    French and German Language.
  • Skill Keywords Extraction Improved for Polish and Spanish Language.
9 June 2020

Latest Updates 

  • Improvement Made in Employer Name and Location Extraction

  • Job Profile Extraction Improved

  • Accuracy of PDF Resume Files Improved

  • French Job Period Extraction Improved

  • Improvement Made in Education and Experience Period for Spanish

  • Spanish Employer Name Extraction Improved

2 June 2020

Latest Update

  • Improvement Made in Polish Resume Education Extraction

  • Improvement did in Spanish Resume for PDF Conversion

  • Personal Information Improved for Spanish Resume

  • Spanish and Polish Language Jobperiods Extraction Improved

  • Job Profile Extraction Improved for Polish and Spanish Language

  • Employername Extraction Improved for the Spanish Language

20 May 2020

Latest Update

  • Deep Learning Models Introduced for the Polish Language for Improved Entities Extractions.
  • Polish Job Period Extraction Improved.
  • Improved Performance Optimization
  • Personal Information Improvement made for the Polish Language.
4 May 2020

Latest Updates

  • Improvement made in Job Period Extraction for Polish Resumes.
  • Improvement made in Extracting Job Profiles if Written in the Same Line i.e Programmer 2014-2016, Sr. Programmer 2016-2018.
  • Improvement made in Employer Name, Job Profile Extraction for Polish Resumes.
22 Apr 2020

                       

Latest Updates

  • Job Profile and Employer Extraction for French, Italian, Turkish Resumes improved.
  • Job Period Extraction improved for French, Italian, Turkish Resumes.
  • Europass Multilingual Parsing Improved.
  • Personal Information like Name and Address for Multilingual Resumes improved.
 
 
 
15 Apr 2020

Latest Updates

  • Improvement made in LinkedIn Resume Parsing. 
  • Improvement made in Normalize Degree.
  • Improvement made in Degree's Specializations.
18 Mar 2020
Latest Updates

 

  • Improvement made in Multi-line candidate name extraction.
  • Improvement made in German and French employers and job profile extraction.
9 Mar 2020

Latest Updates

  • Multilingual LinkedIn Resume Parsing Improved.
  • Job Period Extraction improved for Russian Resumes.
21 Feb 2020

Latest Updates

  • Added new keywords for dates in french Aujourd’hui for fetching dates like "Nov 2016. – Aujourd’hui".
  • Added new splitters like "à fin, À" for dates like Juin 2018 à fin Juillet 2019, AVRIL 2018 À JUIN 2018
  • Experience description filed improved in the segregation of experience for LinkedIn resumes
  • French Degree date improvement made for e.g Septembre 2017 À November 2019
  • Improvement made in Linkedin format name extraction.
  •  Phone format setting for 9 digit phone improved w.r.t resume language e.g. when the location is not defined in France resume and phone no. is of 9 digit phone number is considered as of France.
  • Improvement made in Preserve brackets() in the final JSON output. e.g.   Ass.Engineer (Engineering Department)
  • Improvement made for address city extraction when city name with - e.g. city name like Noisy-le-Grand
10 Feb 2020

 

Latest Updates
  • Improvement made in German Language Dates Extraction
  • Improvement made in For Personal Information for Optional Fields
  • Improvement made in For Special Character (Breaking the Language JSON)
04 Feb 2020
Latest Updates 
  • Improvement made in Parallel Format PDF Text conversion.
  • Improvement made in Degree date format enhanced for dates like '04-16-09 or 05-11-2007'.
  • Improvement made in Name Extraction from new formats eg. name like MS Manoj Singh.
  • Improvement made in Experience accuracy.
  •  Improvement made in Education accuracy.

  • Improvement made in Bugfixing in normalize period

28 Jan 2020

Latest Updates 

  • Improvement made in Education date fixing for Portuguese resume.
  • Improvement made in Adding new splitters for dates i.e (á,até)for Portuguese resume.
  • Improvement made in Phone formatting if the phone is written with ISD code. eg (48) 98154-4522.
  • Improvement made in Special char([&:;\\(\\)']) issue fixing in Job Description.
  • Improvement made in UAE phone formate fixed. 
 
11 Dec 2019

Latest Updates 

  • Improvement made in fetching city, state, country if defined in multiline in the experience like Engineer September 2011 - September 2014
    Hyderabad
    India
  • Improvement made in fetching country name if mentioned without city and state name like HAL Edgewood Technologies
    Reliability Engineer November 2008 - May 2010
    India
09 Nov 2019

Latest Updates 

  • Footer data extraction from pdf files improved.
  • Improvement made in Passport number extraction.
  • Linkedin's new format name extraction improved.
  • Phone number and formatted phone number extraction improved
  • Education data extraction improved.
  • Nested employer extraction improved e.g. if candidate defined employer like :

    Brother International Corporation, Bridgewater, NJ
    Channel Manager (Central America and Caribbean)        June 2017 – Present
    Distribution Sales Account Representative                     April 2016 - May 2017
    Key Account Representative                                            July 2015 - March 2016

    RChilli will return it like :

    1 Employer - Brother International Corporation
        Job Title - Channel Manager (Central America and Caribbean)
        Job period- June 2017 – Present
        Job Location - Bridgewater, NJ


    2 Employer - Brother International Corporation
        Job Title - Distribution Sales Account Representative
        Job period- April 2016 - May 2017
        Job Location - Bridgewater, NJ


    3  Employer - Brother International Corporation
        Job Title - Key Account Representative
        Job period- July 2015 - March 2016
        Job Location - Bridgewater, NJ

    Another Example: -

HANESBRANDS, INC., Winston Salem, NC          2007 - 2019
Senior Manager, eCommerce                             2012 –2019
Search Engine Marketing Manager                     2007 – 2011


RChilli will return it like :


1   Employer - HANESBRANDS, INC.
    Job Title - Senior Manager, eCommerce
    Job period- 2012 –2019
   Job location- Winston Salem, NC
   

2  Employer - HANESBRANDS, INC.
    Job Title - Search Engine Marketing Manager
    Job period- 2007 – 2011
   Job location- Winston Salem, NC
21 Oct 2019

Latest Updates  

  • Fixing of slashes (//) coming at the end of job profiles.
  • Start extracting job period if written like April 2010-September in one line and remaining period 2014 in the next line.
20 Sep 2019

Latest Updates 

  • Improvement made on Summary block extraction.
  • Improvement made in the Formatted phone number.
11 Sep 2019

Latest Updates 

  • Email extraction improvement is done for emails having a symbol (+) like dev+rchilli@gmail.com
  •  Passport number extraction improved.
19 Aug 2019

Latest Updates  

  • Projects mapping is defined without a period with employer defined without job period
  • Passport Number fixing
26 July 2019

Latest Updates 

  • Considering job periods fixing is done even if candidates write them in reverse order, i.e. (Aug 2018 -Mar 2014).
  • Address fixing is done for LinkedIn new format resumes.
19 July 2019

Latest Updates 

  • Years will be written as "April 2010-Aug" and 2011 in the next line.
  • Skill keywords added for oil and petroleum industries for category/subcategory search.
  • Tagging issue fixed in PDF pages.
15 July 2019

Latest Updates 

  • Issue fixed for month extraction if candidate defines only the year in a case where end year is the same as the current year (i.e., 2010-2019 ) here end year '2019' same as the present year '2019.'
  • PDF issue fixed for multiple '&'.
  • Tagging issue fixed for two simultaneous institutes.
04 July 2019

Latest Updates 

  • Improvement made in Turkish address extraction.
  • Improvement has done in Turkish personal information extraction improvement.
  • Pdf conversion improvement for personal information written in the column
  • Started parsing job periods like 2016-6/2017
20 June 2019

Latest Updates 

  • Emails and other personal information issues improved for document conversion.
  • Degree/Subject extraction optimization done.
  • The location issue fixed if the employer contains a location name.

 

05 June 2019

Latest Updates 

  • Employer City, State extraction improved when written without the comma.
  • Ignoring duplicate subjects from degree extraction.
  • Improvement made for incomplete subject and degree extraction.
24 May 2019

Latest Updates 

  • Improved USA address city extraction.
  • Improved algorithm for Degree Name, University Name, Employer, Job profile extraction.
  • Improved country name extraction for the employer along with it's ISO Code.
  • Starts extraction job periods defined in mixed format e.g 24/2/2014 to December 2014
08 May 2019

Latest Updates 

  • Amendments made in a multilingual common word for name extraction.
  • Amendments have done to combine multiple section headings into one heading like (Experience and training etc.). 
  • Danish resume fixed for address and name.
  • PDF header issue fixed.
02 May 2019

Latest Updates 

  • Single-word organization name improved when given with location detection.
  • Singapore address extraction improved.
  • Improved Parallel pdf parsing.
  • Fixed Date pattern like (May-18 till).
24 April 2019

Latest Updates 

  • Address extraction improved for better performance.
  • Improved date fetching patterns when dates written with tabs.
  • Improvements made for XML breaking in the education section under version 6.0.
  • Improvement made to ignore a single subject return when a degree or institute name is not available.
16 April 2019

Latest Updates 

 

RCHILLI RESUME PARSER API

1. Amendments have done in API to fetch experience portion if it comes under the skill or education section while parsing the data.

Amendments have done for the scenarios wherein:- 
* Experience headers are not defined by the candidate in the CV.
* PDF resumes with parallel formats.

If candidates CV defines some sections (Experience, Education, Skill, etc.) 
on the left side and some sections on the right side. Earlier, the parser was not fetching experience accurately in such a scenario.
Now we have made our parser to fetch the same across the CV's

2. Amendment did in the database for employer city issues while parsing.
 
 
04 April 2019

Latest Updates 

 

RCHILLI RESUME PARSER API

  1. Total Experience in year field has been fixed, previously date patterns like '09-4-2007' or '09/4/2007' were not properly calculated. Now it's being calculated perfectly based on the example mentioned.
  2. Improvement in "LanguageKnown" field extraction from resume.
     
RCHILLI RESUME PARSER API
 
Updated the displaying of companies name having "&" in them e.g "L & T".
 
 
28 March 2019

Latest Updates 

 

RCHILLI RESUME PARSER API

1. We have started considering job periods which are not properly mentioned by the candidates in resume neither in ascending order nor in descending order i.e (2001-2004, 2007-2008, 2005-2006).
 
2. We have implemented a check to return the highest total experience after comparing total experience mentioned in an entire cv with the help of "Keywords" (e.g years of experience, Total experience of e.tc) and calculated from job periods calculation.
 
 

RCHILLI RESUME PARSER API

Periods overlapping issues fixed i.e if the candidate has defined periods like 2001-2004 and 2004-2006, for such cases now we have started calculating experiences completely from the start of 2001 to the end of 2004 and then from 2005 till 2006.

22 March 2019

Latest Updates 

RCHILLI RESUME PARSER API

Improved Date Pattern Recognition: 
We started picking up following date patterns in resumes:
 
1) MM-YYYY-dd - MM-dd-YYYY
2) MM/dd/YYYY - dd/MM/YYYY
 
We have also implemented a new algorithm to fetch a single year written in four digits like (2008, 2010, 1998, etc.) and ignore other 4 digits numbers which are not year actually in resume.
15 Feb 2019

Latest Updates 

 

RCHILLI RESUME PARSER API

  • City/State Optimization for the Employer and Performance Enhancement.
  •  Console Log Print Control from Config for Debug and Production.
  • Gap Period Correction did for Overlapped Periods.
  • Fields Accuracy Improved w.r.t Resume Fixing.
  • Performance Improved in Optimizing Regex and Queries.
  • Improved made in PDF Conversion, now we will have Fewer PDF Conversion Issues Encounter.
  

RCHILLI RESUME PARSER API

  • Experience Description Correction.
  • Correction made in Personal Information.
  • Passing year of Education Degree Corrected.
  • Employer Name, City/State Corrected.
  • Fixed Small Letters Tagging of the Company Name, Profiles having to work as, working with.
  • Pick Emails having Spacing with '@'
  • Change in percentage logic.
  • Fixed Conflict with the dates.


11 Jan 2019

Latest Updates 

 

RCHILLI RESUME PARSER API

  • Case-sensitivity issue fixed for University, Institute, Degree, and Employer.
  • Added cities/Places found in Singapore
  • Address Extraction Logic Implemented for Singapore Places.
  • Remove Job-Profile only when Adjoining with the Name.
  • Changes made in Parsing Logs
  • Formatted Date for Education Changed.
  • Block Extraction made without a Section Heading.
  • Multi-Language Date Format.
  • Multiline Name Extraction Using Common Last Names.
  • Employer State Extraction.
 

RCHILLI RESUME PARSER API

  • Correction Made in UK Zip Code.
  • Phone Number Correction for French
  • HTML File Parsing Issue Fixed when Nobody Tag in HTML
  • Address Fixed for USA States when only City and State is Written.
  • Education Period Fixed.
  • Academy Extraction Issue Fixed.

 
22 Nov 2018

Latest Updates 

RCHILLI RESUME PARSER API

Enhanced the algorithm to extract Time period in Education Section.

 

RCHILLI RESUME PARSER API

Linked the Date Format(global setting) with Education section, 
so that customer requested date format for StartDate and EndDate of candidate qualification is parsed.

 

12 Nov 2018

Latest Updates 

RCHILLI RESUME PARSER API

 Personal Info:

The Algorithm Updation has done to Fetch Address from LinkedIn Resume Format.

 Experience Section:

The algorithm changed to Ignore the City/State Name before Employer Name.

 
02 NOV 2018

Latest Updates 

RCHILLI RESUME PARSER API

Updated Name parsing algorithm for doc type of resume:

Changes made in technique to extract the name from a resume(reparse in PDF format).

Accuracy Improved in Experience Block:

Change in the algorithm to Parse Employer Name and Job Profile Experience Block with Special Characters(i.e. @)

RCHILLI RESUME PARSER API

Updated Name Parsing Algorithm for LinkedIn Format Resume:

Now you will get Candidate's Name Parsed from Resumes in the LinkedIn Format.

24 OCT 2018

Latest Updates 

RCHILLI RESUME PARSER API

Updated Date Format for Job period:

Now in Date Format including the keyword 'since', e.g. since Feb 1996 - Jan 2002, we'll pick end date as well.

Addition in Job Profile Logic:

Now if multiple job profiles are mentioned together they'll be parsed together under one experience.

Changes in Education section algo:

Changes made in exception handling of issues and fetching of data from education.

18 OCT 2018

Latest Updates 

RCHILLI RESUME PARSER API

Experience Section Update:

Our expert team has enhanced the login to fetch:

  • Experience Summary, Skills from Experience Block.
  • Employer Name, Job Profile from Jumbled Up Work Experience with only One Job Period.

Addition in the Job Period formats:

Now you can Fetch JobPeriods, mentioned at the beginning of Experience Block, without Range like March 2004, 02/2014.

For Instance: Project Manager, RChilli Inc., March 2010

Employer Name Changes:

We proudly Announce Addition in the Formats of Names Parsed for Organization Name,

i.e. the First Letter of Name is Small followed by Capital Letter and so on.

For Instance: iReality

03 OCT 2018

Latest Updates 

RCHILLI JD PARSER API

  • Improvement made in About Organization and Responsibility
  • Now our Parser Discards 'About Organization' Section for Skill Extraction

RCHILLI RESUME PARSER API


Qualification Section :

  • Updated the Algorithm to Extract the Candidate's Education[Degree in Subject]
  • Updated the Fetching of Education Degree with Location
  • Improved the Process to Fetch Country Details
  • Upgraded our Parser for the Chinese Language
  • Changes in the Parser for Extraction of Address and Website from English-like Resumes

NEW FEATURE

RCHILLI JD PARSER API

New Feature for Limited Users: Bulk Parsing of JDs and Normalization of Output.

 

RCHILLI JD PARSER API

Updated the Parser Algorithm for our Valuable Customers.


RCHILLI RESUME PARSER API

Updated the Parsing Logic with Special Characters.

Facet Resume Parsing Updated for our Valuable customer.

Updation of API for Escape Character and Candidate Image Extension.

04 SEPT 2018

Latest Updates 

RCHILLI RESUME PARSER API

Updated Experience Section: Updation made in the location extraction as listed in the resume.

 

NEW FEATURE

RCHILLI RESUME PARSER API

Added New Response(output) Parameter: Now we Extract CustomId for our Valuable Customers [Version: 6.0.2.1002]

The Recent Addition of Feature: Extract Address Location from Specific Resumes like LinkedIn Resume.