Prominent Research Directions in NLP Alexander Panchenko Assistant - - PowerPoint PPT Presentation

▶

Sep 13, 2023 26 likes •279 views

Prominent Research Directions in NLP Alexander Panchenko Assistant Professor for NLP About myself: a decade of fun R&D in NLP 2002-2008: Bauman Moscow State Technical University , Engineer in Information Systems, MOSCOW 2008: Xerox

SLIDE 1

Prominent Research Directions in NLP

Alexander Panchenko Assistant Professor for NLP

SLIDE 2

About myself: a decade of fun R&D in NLP

2002-2008: Bauman Moscow State

Technical University, Engineer in Information Systems, MOSCOW

2008: Xerox Research Centre Europe,

Research Intern, FRANCE

2009-2013: Université catholique de

Louvain, PhD in Computational Linguistics, BELGIUM

2013-2015: Startup in SNA, Research

Engineer in NLP , MOSCOW

2015-2017: TU Darmstadt, Postdoc in

NLP , GERMANY

2017-2019: University of Hamburg,

Postdoc in NLP , GERMANY

SLIDE 3

About myself: a decade of fun R&D in NLP

Publications in int’l conferences & journals:
ACL
EMNLP
EACL
ECIR
NLE
Best papers @ ReprLearn4NLP

, SemEval’16

Editor and co-chair:
Cambridge Natural Language Engineering (NLE)
Springer LNCS/CCIS: AIST conf.
PC:
ACL, NAACL, EMNLP

, …

SLIDE 4

About myself: my expertise and past/present research foci

Lexical Semantics
Semantic similarity
Word sense

disambiguation

Word/sense embedding
Taxonomy induction,
Frame induction, …
Argument mining
Graph clustering

SLIDE 5

About myself: my expertise and past/present research foci

Lexical Semantics
Semantic similarity
Word sense

disambiguation

Word/sense embedding
Taxonomy induction,
Frame induction, …
Argument mining
Graph clustering

SLIDE 6

Past projects: PhD thesis (2013)

Semantic relatedness
Using large text collections to

learn statistical models of distributional semantics

… with applications to short

text categorization and search.

… EU project iCOP for

categorization of texts.

http://panchenko.me/papers/

thesis.pdf

SLIDE 7

Past projects: PhD thesis (2013)

Semantic relatedness
Using large text collections to

learn statistical models of distributional semantics

… with applications to short

text categorization and search.

… EU project iCOP for

categorization of texts.

http://panchenko.me/papers/

thesis.pdf

SLIDE 8

Past projects: ELIS-IT (2013)

Expertise Localization from

Informal Sources & Information Technologies

Retrieval of skills from text (e.g.

set of corporate documents)

http://cental.fltr.ucl.ac.be/

projects/elisit/index_EN.html

SLIDE 9

Past projects: NLP for Social Network Analysis (2014)

Based on a start-up company

specializing on SNA

Mining Facebook and

VKontakte social networks.

Analysis of posts, groups,

comments, …

… with respect to sentiment,

topic, gender, age, etc.

… mostly for doing user

segmentation and targeting.

SLIDE 10

Past projects: NLP for Social Network Analysis (2014)

Based on a start-up company

specializing on SNA

Mining Facebook and

VKontakte social networks.

Analysis of posts, groups,

comments, …

… with respect to sentiment,

topic, gender, age, etc.

… mostly for doing user

segmentation and targeting.

SLIDE 11

Past projects: NLP for Social Network Analysis (2014)

Based on a start-up company

specializing on SNA

Mining Facebook and

VKontakte social networks.

Analysis of posts, groups,

comments, …

… with respect to sentiment,

topic, gender, age, etc.

… mostly for doing user

segmentation and targeting.

SLIDE 12

Past projects: NLP for Social Network Analysis (2014)

Based on a start-up company

specializing on SNA

Mining Facebook and

VKontakte social networks.

Analysis of posts, groups,

comments, …

… with respect to sentiment,

topic, gender, age, etc.

… mostly for doing user

segmentation and targeting.

… but also some linguistic

studies.

SLIDE 13

Past projects: new/s/leak (2016)

Information extraction

and interactive visualization of textual datasets for investigative data-driven journalism and eDiscovery

Data journalism
http://www.newsleak.io

SLIDE 14

Active research: Computational Semantics

Word sense

disambiguation and induction

Entity Linking
Integration of knowledge

bases into neural networks

Frame semantics

SLIDE 15

Active research: Computational Semantics

Word sense

disambiguation and induction

Entity Linking
Integration of knowledge

bases into neural networks

Frame semantics

SLIDE 16

Active research: Computational Semantics

Word sense

disambiguation and induction

Entity Linking
Integration of knowledge

bases into neural networks

Frame semantics

SLIDE 17

Active research: Computational Semantics

Word sense

disambiguation and induction

Entity Linking
Integration of knowledge

bases into neural networks

Frame semantics

SLIDE 18

Active research: (Comparative) Argument Mining

Sentiment analysis ++
… not only opinions but

also objective arguments.

… from text only.
Retrieve pros and cons

to make some informed decisions.

SLIDE 19

Active research: (Comparative) Argument Mining

Sentiment analysis ++
… not only opinions but

also objective arguments.

… from text only.
Retrieve pros and cons

to make some informed decisions.

SLIDE 20

Active research: (Comparative) Argument Mining

Sentiment analysis ++
… not only opinions but

also objective arguments.

… from text only.
Retrieve pros and cons

to make some informed decisions.

http://ltdemos.informatik.uni-hamburg.de/cam/

SLIDE 21

Active research: (Comparative) Argument Mining

Sentiment analysis ++
… not only opinions but

also objective arguments.

… from text only.
Retrieve pros and cons

to make some informed decisions.

http://ltdemos.informatik.uni-hamburg.de/cam/

SLIDE 22

Active research: (Comparative) Argument Mining

SLIDE 23

Active research: (Comparative) Argument Mining

SLIDE 24

Active research: (Comparative) Argument Mining

SLIDE 25