Errata to the Second Edition

A list of errata to the first edition is available here.

Chapter 5

Page 137, Unix script. Replace the enumerated list with:

tr -cs ’A-Za-z’ ’\n’ <input_file > token_file

Tokenize the input and create a file with the unigrams.

```
tail +2 < token_file > next_token_file
```
Create a second unigram file starting at the second word of the first tokenized file (+2).
```
paste token_file next_token_file > bigrams
```
Merge the lines (the tokens) pairwise. Each line contains the words at index i and i + 1 separated with a tabulation.
And we count the bigrams as in the previous script.

Page 161, in the Perl program, replace the two occurrences of:

log($p) +

with

log($p) -

Page 193, Figure 6.9. Replace the figure with

Corrected figure [tikz].

Page 325, Figure 11.2. Replace the rightmost NP₂ with NP₁. Corrected figure [tikz].

Table 13.12, page 435. The analysis is for the sentence The waiter brought the meal and not The waiter brought a meal

Table 13.12, page 435. Sitxth row from bottom. Change (VBD, NN) to (VBD, DT)

Page 506, Figure 15.13. Replace the two node labels below the noun root node with respectively concrete and abstract. Corrected figure [tikz].

Page 530, Table 16.8, 4th row. Move the word major from the 3rd column to the 4th column.

Institutionen för Datavetenskap 2014. Ansvarig: Pierre Nugues
Last update: Saturday, 19-Dec-2015 16:37:13 CET