1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
|
<?php
class Af_Sort_Bayes extends Plugin {
private $host;
private $filters = array();
private $dbh;
private $score_modifier = 50;
function about() {
return array(1.0,
"Bayesian classifier for tt-rss (WIP)",
"fox");
}
function init($host) {
require_once __DIR__ . "/lib/class.naivebayesian.php";
require_once __DIR__ . "/lib/class.naivebayesianstorage.php";
$this->host = $host;
$this->dbh = Db::get();
$this->init_database();
$host->add_hook($host::HOOK_ARTICLE_FILTER, $this);
$host->add_hook($host::HOOK_PREFS_TAB, $this);
$host->add_hook($host::HOOK_ARTICLE_BUTTON, $this);
}
function trainArticle() {
$article_id = (int) $_REQUEST["article_id"];
$train_up = sql_bool_to_bool($_REQUEST["train_up"]);
$category = $train_up ? "GOOD" : "NEUTRAL";
$nbs = new NaiveBayesianStorage($_SESSION["uid"]);
$nb = new NaiveBayesian($nbs);
$result = $this->dbh->query("SELECT score, guid, title, content FROM ttrss_entries, ttrss_user_entries WHERE ref_id = id AND id = " .
$article_id . " AND owner_uid = " . $_SESSION["uid"]);
if ($this->dbh->num_rows($result) != 0) {
$guid = $this->dbh->fetch_result($result, 0, "guid");
$title = $this->dbh->fetch_result($result, 0, "title");
$content = mb_strtolower($title . " " . strip_tags($this->dbh->fetch_result($result, 0, "content")));
$score = $this->dbh->fetch_result($result, 0, "score");
$this->dbh->query("BEGIN");
if ($nb->untrain($guid, $content)) {
if ($score >= $this->score_modifier) $score -= $this->score_modifier;
}
$nb->train($guid, $nbs->getCategoryByName($category), $content);
if ($category == "GOOD") $score += $this->score_modifier;
$this->dbh->query("UPDATE ttrss_user_entries SET score = '$score' WHERE ref_id = $article_id AND owner_uid = " . $_SESSION["uid"]);
$nb->updateProbabilities();
$this->dbh->query("COMMIT");
}
print "$article_id :: $category";
}
function get_js() {
return file_get_contents(__DIR__ . "/init.js");
}
function hook_article_button($line) {
return "<img src=\"plugins/af_sort_bayes/thumb_up.png\"
style=\"cursor : pointer\" style=\"cursor : pointer\"
onclick=\"bayesTrain(".$line["id"].", true)\"
class='tagsPic' title='".__('+1')."'>" .
"<img src=\"plugins/af_sort_bayes/thumb_down.png\"
style=\"cursor : pointer\" style=\"cursor : pointer\"
onclick=\"bayesTrain(".$line["id"].", false)\"
class='tagsPic' title='".__('-1')."'>";
}
function init_database() {
$prefix = "ttrss_plugin_af_sort_bayes";
// TODO there probably should be a way for plugins to determine their schema version to upgrade tables
/*$this->dbh->query("DROP TABLE IF EXISTS ${prefix}_wordfreqs", false);
$this->dbh->query("DROP TABLE IF EXISTS ${prefix}_references", false);
$this->dbh->query("DROP TABLE IF EXISTS ${prefix}_categories", false);*/
$this->dbh->query("BEGIN");
// PG only for the time being
$this->dbh->query("CREATE TABLE IF NOT EXISTS ${prefix}_categories (
id SERIAL NOT NULL PRIMARY KEY,
category varchar(100) NOT NULL DEFAULT '',
probability DOUBLE PRECISION NOT NULL DEFAULT '0',
owner_uid INTEGER NOT NULL REFERENCES ttrss_users(id) ON DELETE CASCADE,
word_count BIGINT NOT NULL DEFAULT '0')");
$this->dbh->query("CREATE TABLE IF NOT EXISTS ${prefix}_references (
id SERIAL NOT NULL PRIMARY KEY,
document_id VARCHAR(255) NOT NULL,
category_id INTEGER NOT NULL REFERENCES ${prefix}_categories(id) ON DELETE CASCADE,
owner_uid INTEGER NOT NULL REFERENCES ttrss_users(id) ON DELETE CASCADE,
content text NOT NULL)");
$this->dbh->query("CREATE TABLE IF NOT EXISTS ${prefix}_wordfreqs (
word varchar(100) NOT NULL DEFAULT '',
category_id INTEGER NOT NULL REFERENCES ${prefix}_categories(id) ON DELETE CASCADE,
owner_uid INTEGER NOT NULL REFERENCES ttrss_users(id) ON DELETE CASCADE,
count BIGINT NOT NULL DEFAULT '0')");
$owner_uid = @$_SESSION["uid"];
if ($owner_uid) {
$result = $this->dbh->query("SELECT id FROM ${prefix}_categories WHERE owner_uid = $owner_uid LIMIT 1");
if ($this->dbh->num_rows($result) == 0) {
$this->dbh->query("INSERT INTO ${prefix}_categories (category, owner_uid) VALUES ('GOOD', $owner_uid)");
$this->dbh->query("INSERT INTO ${prefix}_categories (category, owner_uid) VALUES ('NEUTRAL', $owner_uid)");
}
}
$this->dbh->query("COMMIT");
}
function hook_prefs_tab($args) {
if ($args != "prefPrefs") return;
print "<div dojoType=\"dijit.layout.AccordionPane\" title=\"".__('af_sort_bayes')."\">";
//
print "</div>";
}
function hook_article_filter($article) {
$owner_uid = $article["owner_uid"];
$nbs = new NaiveBayesianStorage($owner_uid);
$nb = new NaiveBayesian($nbs);
$categories = $nbs->getCategories();
if (count($categories) > 0) {
$count_neutral = 0;
$count_good = 0;
$id_good = 0;
$id_neutral = 0;
foreach ($categories as $id => $cat) {
if ($cat["category"] == "GOOD") {
$id_good = $id;
$count_good += $cat["word_count"];
} else if ($cat["category"] == "NEUTRAL") {
$id_neutral = $id;
$count_neutral += $cat["word_count"];
}
}
$dst_category = $id_neutral;
$bayes_content = mb_strtolower($article["title"] . " " . strip_tags($article["content"]));
if ($count_neutral >= 3000 && $count_good >= 1000) {
// enable automatic categorization
$result = $nb->categorize($bayes_content);
if (count($result) == 2) {
$prob_good = $result[$id_good];
$prob_neutral = $result[$id_neutral];
if ($prob_good > 0.90 && $prob_good > $prob_neutral) {
//$dst_category = $id_good; // should we autofile as good or not? idk
$article["score_modifier"] += $this->score_modifier;
}
}
}
$nb->train($article["guid_hashed"], $dst_category, $bayes_content);
$nb->updateProbabilities();
}
return $article;
}
function api_version() {
return 2;
}
}
?>
|
