diff --git a/Markov_youtube_description.py b/Markov_youtube_description.py
new file mode 100644
index 0000000..8121a1d
--- /dev/null
+++ b/Markov_youtube_description.py
@@ -0,0 +1,81 @@
+""" Text mining project: finds the descriptions of Youtube videos using BeautifulSoup and makes a new description using Markov analyis """ 
+from bs4 import BeautifulSoup
+import requests
+import string
+import random
+import pickle
+import doctest
+
+"""
+Youtube Descriptions
+"""
+
+def yt_description(youtube_id):
+	"""Returns the youtube description of a video.
+	>>> yt_description('KenrpZ1oxfU')
+	["Republican candidate weighs in on Israeli-Palestinian conflict, trade with China, 9/11 Commission and Iraq War on 'Hannity'"]
+	>>> yt_description('SWl3xfSqOIY')
+	["On 'Hannity,' presidential candidate responds to negative ads, talks feud with pope, illegal immigration"]
+	"""
+	printed_description = [] #list that holds Youtube description
+	html = requests.get('https://www.youtube.com/watch?v={}'.format(youtube_id))
+	soup = BeautifulSoup(html.text, "lxml")
+	description = soup.find('p',id= 'eow-description') #finds the description in a YouTube video
+
+	printed_description.append(str(description.get_text())) #converts description to only text and appends it to an empty list
+	return printed_description
+
+def combine_descriptions(video_ids):
+	"""Combines descriptions of different youtube videos into one list
+	>>> combine_descriptions(['KenrpZ1oxfU','SWl3xfSqOIY'])
+	["Republican candidate weighs in on Israeli-Palestinian conflict, trade with China, 9/11 Commission and Iraq War on 'Hannity'", "On 'Hannity,' presidential candidate responds to negative ads, talks feud with pope, illegal immigration"]
+	"""
+	combined_descriptions = [] #list that holds the combined YouTube video descriptions
+	for video_id in video_ids:
+		combined_descriptions += yt_description(video_id) #combines descriptions into one list
+	return combined_descriptions
+
+"""
+Markov analysis
+"""
+
+prefix = ()
+suffix_map = {}
+
+def Markov_file(filename,order):
+	for line in filename:
+		for word in line.strip().split():
+			store_words(word,order)
+
+def store_words(word, order):
+	global prefix
+	if len(prefix)<order:
+		prefix = prefix + (word,)
+		return
+	try:
+		suffix_map[prefix].append(word)
+	except KeyError:
+		suffix_map[prefix] = [word] # makes a prefix if there are none
+
+	prefix = shift(prefix, word)
+
+def random_text(n):
+    start = random.choice(suffix_map.keys()) #chooses random prefix 
+    
+    for i in range(n):
+        suffixes = suffix_map.get(start, None)
+        if suffixes == None: # if we get to  the end of the text, start again
+            random_text(n-i)
+            return
+
+        word = random.choice(suffixes)
+        print word,
+        start = shift(start, word)
+
+def shift(tup,word):
+	return tup[1:] + (word,) #makes a new tuple by removing head and adding word to the end
+
+if __name__ == '__main__':
+	doctest.testmod()
+	Markov_file(combine_descriptions(['eCXSpllybUM','IPkHW0mA8O4','Rhp7oamdVow','jMkDv8O_yoA','se9iaOX8bgc','I4cXVxCrqPc','ckzGe3w-gOI','Oc7DWPGb18k','zwK9iqkLOpc','KenrpZ1oxfU']),2)
+	random_text(50)
\ No newline at end of file
diff --git a/text_mining_writeup.pdf b/text_mining_writeup.pdf
new file mode 100644
index 0000000..22f51b8
Binary files /dev/null and b/text_mining_writeup.pdf differ