Возможность применения анализа настроения пользователей Twitter для повышения точности прогноза движения цен инструментов фондового рынка

Thu May 28 05:19:37 +0000 2015"RT @SexualGif: a night like this would be chill \ud83c\udf0c\ud83c\udf06\u26fa\ufe0f http:\/\/t.co\/K1wi3WxX8N

Thu May 28 05:19:38 +0000 2015"\u201c@KrisVinesGames: @IonnaMcKenzie Looking good \ud83d\ude09\ud83d\udc4d\u201dthank you\u2764\ufe0f

Thu May 28 05:19:38 +0000 2015"RT @NBA_Skits: Lil B reminding the crowd who owns the cooking dance. \nhttp:\/\/t.co\/uYCPMRmZI5

Thu May 28 05:19:38 +0000 2015"RT @NYTHealth: Strong evidence that treating more people with H.I.V. earlier would save more lives http:\/\/t.co\/WWXqOKQcnh

Thu May 28 05:19:38 +0000 2015"Hi @Harry_Styles!\nI love you so much \ud83d\udc98\nMy birthday is on \nJune 3rd, & \nIt'll mean a lot for your,\nCongratulations! \nx22

Thu May 28 05:19:38 +0000 2015"Girls Who Squat Before And After. http:\/\/t.co\/K52O1EoSAm #Workout #diet #healthy #girls #fitness #fispo http:\/\/t.co\/5ifwaALASD

Thu May 28 05:19:38 +0000 2015"#ChoiceTVChemistry Oliver & Felicity \n#ChoiceTVBreakOutStar Emily Bett Rickards & Carlos Valdes #ChoiceSciFiTVShow Arrow \n#TeenChoice 14

Thu May 28 05:19:38 +0000 2015"#NowPlaying \My Ghost\" Glass Pear da Streets Of Love \u266b http:\/\/t.co\/g1QmoTHs6m"

Thu May 28 05:19:38 +0000 2015"Stalkistice Yo

Thu May 28 05:19:38 +0000 2015"RT @femaIes: this just kills me... http:\/\/t.co\/9TznSdRRvH

Thu May 28 05:19:38 +0000 2015"@DMartinez_04 lol I passed all my classes though \ud83d\ude02\ud83d\ude02\ud83d\ude02

Thu May 28 05:19:38 +0000 2015"Translates into 8 out of 10 days of pure profits! http:\/\/t.co\/QCPqAe1qTm #binarytradingoptions

Thu May 28 05:19:38 +0000 2015"76 My #TeenChoice nominee for #ChoiceMusicNextBigThing is Zayn Malik http:\/\/t.co\/e2iBEv1S90

Thu May 28 05:19:38 +0000 2015"RT @LOHANTHONY: bye dana :( keep the east coast warm and fabulous for me. never not all smiles when i'm around this\u2026 https:\/\/t.co\/9m3fNWYL5u

Thu May 28 05:19:38 +0000 2015"Let's be honest, everyone was watching Riley Curry during that press conference. #RileyCurry #toocute

Thu May 28 05:19:38 +0000 2015"I just wanna cuddle



Приложение 2

В приложении подробно описана реализация программы сентиментного анализа твиттов. Программа написана на языке С.

#pragma warning(disable: 4996)

#include <string.h>

#include <stdio.h>

using namespace std;

//////////////////////////главная функция//////////////////////////

int main()

{

int* semant_func(char*);

FILE *fin;

FILE *fout;

char arr[2000];

char date[10], tmp_date[10];

char time[2], tmp_time[2];

char *tweet;

int flag = 0;

int total_freq=0;

float temp;

int flag_time = 0;

int* result;

int result2[8] = { 0, 0, 0, 0, 0, 0, 0, 0 };

int n = 0;

fin = fopen("Sentiment_Input_29_05.txt", "r"); ////входной файл

fout = fopen("Sentiment_Output_29_05.txt", "w"); ////////////выходной файл

while (fgets(arr, 2000, fin) != NULL){

for (int i = 0; i < 10; i++) {

tmp_date[i] = arr[i];

}

for (int j = 0; j < 2; j++){

tmp_time[j] = arr[11 + j];

}

tweet = &arr[31];

if (flag == 0) {

for (int i = 0; i < 10; i++) { date[i] = tmp_date[i]; fprintf(fout, "%c", date[i]); }

fprintf(fout, "\n");

fprintf(fout, "\ntime: ");

for (int j = 0; j < 2; j++){

time[j] = tmp_time[j];

fprintf(fout, "%c", time[j]);

}

//fprintf(fout, "\n");

flag = 10;

}

for (int i = 0; i < 10; i++){

if (date[i] != tmp_date[i]){

date[i] = tmp_date[i];

fprintf(fout, "%c", date[i]);

for (int i = 0; i < 8; i++){

result2[i] = 0;

}

}

}

result = semant_func(tweet);//////вызов функции ///сентиментного анализа

for (int i = 0; i < 8; i++){

result2[i] = result2[i] + result[i];

}///-----

for (int j = 0; j < 2; j++){

if (time[j] != tmp_time[j]){

time[j] = tmp_time[j];

flag_time = 1;

}

}

if ((flag_time == 1)){

fprintf(fout, "\nfrequencies: ");/////////вывод частот

for (int y = 0; y < 8; y++) {

fprintf(fout, " %d ", result2[y]);

total_freq = total_freq + result2[y];

}

fprintf(fout, "\nnormalized_frequencies: ");

//вывод нормализованных частот

if (total_freq != 0){

for (int y = 0; y < 8; y++) {

temp = result2[y];

temp = (temp / total_freq);

fprintf(fout, " %f ", temp);

}

}

fprintf(fout, "\ntime: ");

for (int j = 0; j < 2; j++){

fprintf(fout, "%c", time[j]);

}

for (int y = 0; y < 8; y++) {

total_freq = result2[y] = 0;

}

flag_time = 0;

flag = 1;

}

}

fprintf(fout, "\nfrequencies: ");

for (int y = 0; y < 8; y++) {

fprintf(fout, " %d ", result2[y]);

total_freq = total_freq + result2[y];

}

fprintf(fout, "\nnormalized_frequencies: ");

if (total_freq != 0){

for (int y = 0; y < 8; y++) {

temp = result2[y];

temp = (temp / total_freq);

fprintf(fout, " %f ", temp);

}

}

fclose(fin);

fclose(fout);

return 0;

}

//////////////////функция сентиментного анализа//////////////////

int* semant_func(char* message){

int freq_arr[8] = { 0,0,0,0,0,0,0,0 };

char* emotions8[8] = { "happy", "loving", "calm", "energetic", "fearful/anxious", "angry", "tired", "sad"};

//freq_arr[0] = "happy"

char* happy[20] = {"happy", "lovely", "felicitous", "glad", "well - chosen", "blessed", "blissful", "bright", "golden", "halcyon", "prosperous", "laughing", "riant", "fortunate", "willing","adorable", "endearing", "beautiful", "lovable", "loveable"};

//freq_arr[1] = "loving"

char* loving [31] = { "loving", "caring", "adoring", "doting", "fond", "affectionate", "fond", "lovesome", "tender", "amative", "amorous", "amatory", "romantic", "attached", "captivated", "charmed", "enamored", "infatuated", "in love", "potty", "smitten", "soft on", "taken with", "idolatrous", "loverlike", "loverly", "overfond", "tenderhearted", "touchy - feely", "uxorious","compassionate"};

//freq_arr[2] = "calm"

char* calm [19]= { "calm", "content", "unagitated", "serene", "tranquil", "composed", "placid", "quiet", "still", "smooth", "unruffled", "settled", "windless",

"contented", "complacent", "self - satisfied", "self - complacent", "satisfied", "smug"};

//freq_arr[3] = "energetic"

char* energetic [47]= { "active", "peppy", "combat - ready", "fighting", "participating", "alive", "dynamic", "progressive", "operational", "activist", "activistic", "hands - on", "proactive", "involved", "about", "astir", "acrobatic", "athletic", "gymnastic", "agile", "nimble", "quick", "spry", "hot", "hyperactive", "overactive", "on the go", "sporty", "activated", "counteractive", "surface - active", "brisk", "bustling", "busy", "going", "open", "springy", "existent", "existing", "alive", "live", "eruptive", "activated",

"bouncing", "bouncy", "spirited", "zippy"};

//freq_arr[4] = "fearful/anxious"

char* fearful[23] = { "jittery", "nervous", "edgy", "high - strung", "highly strung", "jumpy", "nervy", "overstrung", "restive", "uptight", "tense",

"anxious", "queasy", "uneasy", "unquiet", "neural", "aflutter", "skittish", "flighty", "spooky", "troubled", "excited", "excitable"};

//freq_arr[5] = "angry"

char* angry[14] = { "grouchy", "fed up", "crabbed", "crabby", "cross", "fussy", "grumpy", "bad - tempered", "ill - tempered", "ill - natured",

"disgusted", "sick", "tired of", "displeased"};

//freq_arr[6] = "tired"

char* tired [60]= { "tired", "drowsy", "banal", "commonplace", "hackneyed", "old - hat", "shopworn", "stock", "threadbare", "timeworn", "trite", "well - worn", "all in", "beat", "bushed", "dead", "aweary", "weary", "bleary", "blear", "bleary - eyed", "blear - eyed", "bored", "world - weary", "burned - out", "burnt - out", "careworn", "drawn", "haggard", "raddled", "worn", "drooping", "flagging", "exhausted", "dog - tired", "fagged", "fatigued", "played out", "spent", "washed - out", "worn - out", "worn out", "gone", "footsore", "jaded", "wearied", "knackered", "drained", "ragged", "travel - worn", "unrefreshed", "unrested", "whacked", "unoriginal",

"drowsing", "dozy", "oscitant", "yawning", "asleep", "inattentive"};

//freq_arr[6] = "sad"

char* sad [45]= { "gloomy", "sad", "glooming", "gloomful", "sulky", "grim", "blue", "depressed", "dispirited", "down", "downcast", "downhearted", "low", "low-spirited", "blue", "dark", "dingy", "disconsolate", "dismal", "grim", "sorry", "drab", "drear", "dreary", "dejected", "depressing", "cheerless", "uncheerful",

"deplorable", "distressing", "lamentable", "pitiful", "sorry", "bittersweet", "doleful", "mournful", "heavyhearted", "pensive", "wistful", "tragic", "tragical", "tragicomic", "tragicomical", "sorrowful", "bad"};

int length_arr[8] = {20,31,19,47,23,14,60,45};

for (int i = 0; i < length_arr[0]; i++){

if (strstr(message, happy[i]) != NULL){

freq_arr[0] = freq_arr[0] + 1;

}

}

for (int i = 0; i < length_arr[1]; i++){

if (strstr(message, loving[i]) != NULL){

freq_arr[1] = freq_arr[1] + 1;

}

}

for (int i = 0; i < length_arr[2]; i++){

if (strstr(message, calm[i]) != NULL){

freq_arr[2] = freq_arr[2] + 1;

}

}

for (int i = 0; i < length_arr[3]; i++){

if (strstr(message, energetic[i]) != NULL){

freq_arr[3] = freq_arr[3] + 1;

}

}

for (int i = 0; i < length_arr[4]; i++){

if (strstr(message, fearful[i]) != NULL){

freq_arr[4] = freq_arr[4] + 1;

}

}

for (int i = 0; i < length_arr[5]; i++){

if (strstr(message, angry[i]) != NULL){

freq_arr[5] = freq_arr[5] + 1;

}

}

for (int i = 0; i < length_arr[6]; i++){

if (strstr(message, tired[i]) != NULL){

freq_arr[6] = freq_arr[6] + 1;

}

}

for (int i = 0; i < length_arr[7]; i++){

if (strstr(message, sad[i]) != NULL){

freq_arr[7] = freq_arr[7] + 1;

}

}

return freq_arr;

}

Размещено на Allbest.ru

...