Skip to content

Commit

Permalink
layout
Browse files Browse the repository at this point in the history
  • Loading branch information
christofs committed Jul 18, 2017
1 parent c03acb1 commit 67a6fee
Showing 1 changed file with 9 additions and 8 deletions.
17 changes: 9 additions & 8 deletions index.html
Original file line number Diff line number Diff line change
Expand Up @@ -143,7 +143,7 @@
--
## Naive Baseline:<br/>ratio of relative frequencies (rrf)
<br/>
* **rrf<sub>i</sub> = rf<sub>i</sub>(T) / rf<sub>i</sub>(C)**
**rrf<sub>i</sub> = rf<sub>i</sub>(T) / rf<sub>i</sub>(C)**
<br/>
<br/>
* T = Target partition
Expand Down Expand Up @@ -185,9 +185,9 @@


--
## Empirical evaluation [REVISE]
## Empirical evaluation
<br/>
<a href="img/testing-matilda_Lijffijt.png"><img height="250" src="img/testing-matilda_Lijffijt.png"></img></a>
<a href="img/testing-matilda_Lijffijt.png"><img height="300" src="img/testing-matilda_Lijffijt.png"></img></a>
* Lijffit et al. 2014 using BNC
* "Matilda" in male/female authors
* Challenge: "spiky" distribution
Expand All @@ -204,7 +204,7 @@
* John Burrows 2007: proposes Zeta (follow-up to Delta) <!-- .element: class="fragment" data-fragment-index="1" -->
* Hugh Craig 2009: slightly modified version <!-- .element: class="fragment" data-fragment-index="2" -->
* David Hoover 2010: Excel implementation <!-- .element: class="fragment" data-fragment-index="3" -->
* Maciej Eder 2013: R implementation (in stylo) <!-- .element: class="fragment" data-fragment-index="4" -->
* Maciej Eder 2013: R implementation (in stylo: 'oppose()') <!-- .element: class="fragment" data-fragment-index="4" -->
<br/>
<br/>
<p>(see [references](#/8/2))</p> <!-- .element: class="fragment" data-fragment-index="5" -->
Expand All @@ -224,7 +224,7 @@
--
## Zeta, 2: document frequencies
<br/>
* **df<sub>i</sub>(T) and df<sub>i</sub>(C)**
**df<sub>i</sub>(T) and df<sub>i</sub>(C)**
<br/>
<br/>
* df<sub>i</sub> = number of documents <br/>in partition that contain word i
Expand All @@ -233,8 +233,9 @@
--
## Zeta, 3: document proportions
<br/>
* **dp<sub>i</sub>(T) = df<sub>i</sub>(T) / n(T)**
* **dp<sub>i</sub>(C) = df<sub>i</sub>(C) / n(C)**
**dp<sub>i</sub>(T) = df<sub>i</sub>(T) / n(T)**

**dp<sub>i</sub>(C) = df<sub>i</sub>(C) / n(C)**
<br/>
<br/>
* dp<sub>i</sub> = document proportion, i.e. proportion <br/>of documents that contain word i
Expand All @@ -244,7 +245,7 @@
--
## Zeta, 4: subtraction
<br/>
* **Zeta<sub>i</sub> = dp<sub>i</sub>(T) - dp<sub>i</sub>(C)**
**Zeta<sub>i</sub> = dp<sub>i</sub>(T) - dp<sub>i</sub>(C)**
<br/>
<br/>
* Not a ratio, but simple subtraction!
Expand Down

0 comments on commit 67a6fee

Please sign in to comment.