1. eHS AI component roadmap: Step I: prototype with fuzzy matching
Maria BIRYUKOV University of Luxembourg

2. Premises
Fuzzy matching step is intended to help user in finding controlled vocabulary (CV) terms, corresponding to his/her terminology.
Input: User-provided term (one or multiword expression)
Output: |N| best corresponding terms, along with their unique identifiers, from standard vocabulary. |N| can be specified by the user.

3. Overall view and Timeline
AI mechanisms for standardizing vocabulary based on eMDR
User Input [Unmapped Variable name or Value]
User output
[ranked list of candidate controlled terms]
Fuzzy string matching
Contextual disambiguation
Semantic inference
eMDR [Synonyms mapping dictionary]
Timeline:
15:10:2015
15:12:2015
31:01:2016
Prototype with Fussy matching
Contextual disambiguation
Semantic inference

4. Achievements DONE Timeline 15:10:2015 Prototype with Fussy matching
15:12:2015
Contextual disambiguation
Timeline
31:01:2016
Semantic inference

5. Resources Thematic dictionaries, ontologies, databases
In absence of eMDR, Entrez gene data base is used as a resource to test and demonstrate functionality of the prototype
Locally created and regularly updated Question/Answers Data Base (QADB) which stores user queries along with the answers, selected as appropriate by users .
Eventually stored in eMDR when eMDR is ready

6. Procedure User is prompted to :
1. Introduce his/her query
2. Specify method he/she would like to use in order to find corresponding standard terms
3. Specify the max number of candidates to return (matches to show)
4. Specify the similarity threshold
Step (1) is obligatory. Steps (2 – 4) are optional. If not provided, default parameters are used.
Steps 2 –> 4 = “regular search”

7. Procedure
Once user has typed in the query (Q), QADB lookup is performed.
If Q is in QADB:
Answers to Q are displayed, from most to less popular
User is prompted to select the answer which corresponds to his/her intention, if there is one
If the Q is answered, user may either introduce new query or quit
If the displayed answers do not satisfy the user, he/she may either proceed for the regular DB search (go to step 2, see previous slide) or quit.
If Q is not in QADB, the procedure continues from step 2 ( see previous slide)

8. Local Resource Maintenance
All the queries are stored along with the information about how frequent they are in “All-Queries Data Base” (AQDB)
Queries for which no answer was found in the resources can be worked off-line and serve for the resource enrichment.

9. Fuzzy String Matching Methods
Three methods for fuzzy string matching are implemented:
‘Gestalt pattern matching’ (1)
Ngram-based cosine similarity (2)
Word-based cosine similarity (3)
(1-3) are appropriate for fuzzy string matching and often produce similar results. However:
(1, 2) better handle spelling mistakes,
(3) is more robust for word order changes or word omission.
We will test the methods with real data and, depending on the results, keep, remove or add methods.

10. Example
For illustration purpose let’s assume that user’s queries are some protein names which he/she would like to map to standard Entrez Gene names and identifiers.
Query 1: ‘steroid hormone receptor’.

11. Search and Results The DB was already searched for that term earlier.
The candidate answers are proposed in the order of their “popularity”: the highest # of votes first
“Other aliases/designations” are alternative spellings of the standard name
“Organism” illustrates the ambiguity
User may select the best option or ‘Nothing’ if no one suits his/her needs
In this example, user opts for answer number 2
User’s choice is recorded and the QA database is updated.

12. QA Database Update
User’s choice is accounted for immediately as suggested by the order of the candidate answers (compare with the previous slide)
If user does not like any of the proposed answers, his/her query can be processed against the whole system database, i.e. ‘regular search’
Note, when user’s query is not found in the QADB, the procedure follows “regular search” path directly

13. Searching Process
Regular search elements: query, string comparison method, max number of suggestions to display, similarity threshold
User may select one, two or three methods
String similarity threshold 0.00 = threshold value will be applied internally depending on the method.

14. Results
Methods are applied one after another
String similarity score
‘Show more’ option
Answers are displayed from the highest to lowest string similarity score
N top-ranked items are shown, N = ‘max matches to display’ (see previous slide)
If ‘show more’ option is chosen, the answers are displayed by batches
‘Best answer’ can be selected from any/all batch (es)
It is not contradictory, as the query accumulates ‘votes’

15. End of the session Continue with the same query
2nd method from the user’s method selection

16. Local resources after user session(s)
Fragment of the “All-Queries Data Base” (AQDB)
Query
Query frequency
Queries are systematically stored and their overall counter is maintained.
It allows for:
Local resource enrichment
Grouping of queries by projects

17. Local resources after user session(s)
Fragment of the “Question/Answers Data Base ” (QADB)
Query
Term ID : Votes
QADB stores query and answers which have been selected by users as most suitable
“Query” is a one or multi-word expression, “answer” is the standard term name and unique identifier
Assuming (potential) ambiguity, many standard IDs may correspond to the same query
Votes = how many times users have selected specific ID for the term.
Votes are accumulated throughout all sessions.
In the example above, “steroid hormone receptor” was mostly selected as “ESRRA” of human (5 times); and equally as “Esrra” of Norway rat, and “esrra” of zebrafish.

18. Next session: already seen query
If this term is searched again, the QADB suggestions will be displayed in following order:

19. Challenges
Need real data and eMDR in order to test the implemented, fuzzy string matching, step.
Adjust according to the results and with respect to the data
Implement next two steps.
Need real data and eMDR
The command line demo will be provided later as an API, once the form of the API is agreed with other WPs.

20. Thanks
To Reinhard Schneider, Wei Gu, Venkata for fruitful discussions and advise To Fabien Chris for valuable comments Thank you for your attention