Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Next revision
Previous revision
public:reimporting_prepositions:start [2017-04-27 09:55] – created Johannes Graënpublic:reimporting_prepositions:start [2023-09-15 20:33] (current) – external edit 127.0.0.1
Line 1: Line 1:
 ====== Crossing the Border Twice: Reimporting Prepositions to Alleviate L1-Specific Transfer Errors ====== ====== Crossing the Border Twice: Reimporting Prepositions to Alleviate L1-Specific Transfer Errors ======
  
-===== Overview =====+This page contains the most relevant verb preposition (VPC) and adjective preposition combinations (APC) for English learners with different L1's from our publication [[http://www.zora.uzh.ch/137099/|Crossing the Border Twice: Reimporting Prepositions to Alleviate L1-Specific Transfer Errors]].  The lists have been filtered such that the raw frequency $f_V$ is required to be at least 100 and //by// is not allowed as preposition $\lambda_p$ due to the frequent use of passive verb forms in [[public:costep:start|source material]].  The same lists with a filter of $f_V \ge 5$ are available [[https://pub.cl.uzh.ch/users/graen/vpc_apc_export/|here]]. 
 + 
 + 
 +===== Specifications ===== 
 + 
 +  * $\lambda_v$ -- the verb (or adjective) lemma 
 +  * $\lambda_p$ -- the corresponding preposition 
 +  * $\lambda_{p^{\prime\prime}}$ -- the backtranslated preposition 
 +  * $f_V$ -- the raw frequency of VPC ($\lambda_v$,$\lambda_p$) 
 +  * $p_V$ -- frequency of VPC relative to $\lambda_v$: $p_V (\lambda_p|\lambda_v)$ 
 +  * $p_{V^{\prime\prime}}$ -- frequency of language-specific VPC + backtranslated preposition relative to VPC + language $\gamma$: $p_{V^{\prime\prime}} (\lambda_{p^{\prime\prime}}|\lambda_v,\lambda_p,\gamma)$ 
 +  * $w_{V^{\prime\prime}}$ -- backtranslation ratio (BTR): $w_{V^{\prime\prime}} (\lambda_{p^{\prime\prime}}|\lambda_v,\lambda_p,\gamma) = \frac{ p_{V^{\prime\prime}} (\lambda_{p^{\prime\prime}}|\lambda_v,\lambda_p,\gamma) }{ p_{V^{\prime\prime}} (\lambda_p|\lambda_v,\lambda_p,\gamma) }$ 
 +  * $\alpha_V$ -- BTR-related weight (higher BTR lead to higher values) 
 +  * $\beta_V$ -- VPC-related weight (higher frequency of VPC leads to higher value) 
 +  * $\omega_V$ -- combined weight used for ranking
  
-===== Data Extraction ===== 
  
 ===== Recommendation Lists ===== ===== Recommendation Lists =====
  
-==== German VPC ====+==== German L1 VPC ====
  
 ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^ ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^
Line 92: Line 105:
  
  
-==== French VPC ====+==== French L1 VPC ====
  
 ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^ ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^
Line 221: Line 234:
  
  
-==== Italian VPC ====+==== Italian L1 VPC ====
  
 ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^ ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^
Line 322: Line 335:
  
  
-==== Polish VPC ====+==== Polish L1 VPC ====
  
 ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^ ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^
Line 426: Line 439:
  
  
-==== Spanish VPC ====+==== Spanish L1 VPC ====
  
 ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^ ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^
Line 613: Line 626:
  
  
-==== German APC ====+==== German L1 APC ====
  
 ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^ ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^
Line 647: Line 660:
  
  
-==== French APC ====+==== French L1 APC ====
  
 ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^ ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^
Line 679: Line 692:
  
  
-==== Italian APC ====+==== Italian L1 APC ====
  
 ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^ ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^
Line 711: Line 724:
  
  
-==== Polish APC ====+==== Polish L1 APC ====
  
 ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^ ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^
Line 743: Line 756:
  
  
-==== Spanish APC ====+==== Spanish L1 APC ====
  
 ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^ ^ ^ $\lambda_v$ ^ $\lambda_p$ ^ $\lambda_{p^{\prime\prime}}$ ^  $f_V$ ^  $p_V$ ^  $p_{V^{\prime\prime}}$ ^  $w_{V^{\prime\prime}}$ ^  $\alpha_V$ ^  $\beta_V$ ^  $\omega_V$ ^

CL Wiki

Institute of Computational Linguistics – University of Zurich