The example below uses the LINQ to XML predecessor System.Xml.XmlDocument simply because I still prefer the regex parsing method. It can easily be adapted to use XDocument instead based on your preference. The end resulting output (IEnumerable) is the same.

code:

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Xml;
using System.Dynamic;
using System.Xml.Linq;

namespace AIS.Common
{
    public class DocParser : IDisposable
    {
        public enum DocTypes
        {
            Xml
        };

        private XmlDocument Document { get; set; }
        //private XDocument Document { get; set; }
        private DocTypes DocType { get; set; }

        //constructor
        /// <summary>
        /// 
        /// </summary>
        /// <param name="dt">type of document, currently only supports xml</param>
        /// <param name="f">filename and path</param>
        public DocParser(DocTypes dt, string fp)
        {
            try
            {
                switch (dt)
                {
                    case DocTypes.Xml:
                        DocType = DocTypes.Xml;

                        Document = new XmlDocument();
                        Document.Load(fp);

                        //Document = XDocument.Load(fp);
                        break;
                    default:
                        throw new Exception("Constructor::invalid or unknown doc type specified");
                }
            }
            catch (Exception ex)
            {
                throw new Exception("Constructor::Error with document load or doc type.",ex);
            }
        }

        public IEnumerable<dynamic> GetElements(string regex)
        {
            return XmlToExpandoObject(regex);
        }

        private IEnumerable<dynamic> XmlToExpandoObject(string regex)
        {
            dynamic eo = new ExpandoObject();
            var dictionary = eo as IDictionary<string, object>;
            XmlNodeList xnl = Document.SelectNodes(regex);
            foreach (XmlNode xn in xnl)
            {
                foreach (XmlAttribute xa in xn.Attributes)
                {
                    dictionary[xa.Name.ToString()] = xa.Value.Trim();
                    //TrySetAttr(xn.LocalName, xn);
                }

                yield return eo;
            }
        }
}
}

sample xml:

<?xml version="1.0" encoding="utf-8" ?>
<robots>
  <robot id="1" age="5" />
  <robot id="2" age="3" />
</people>

usage:

DocParser dp = new DocParser(DocParser.DocTypes.Xml, Server.MapPath("~/xmlfilename.xml")); //can also retrieve xml from external service, etc
        var robots = dp.GetElements("*/robot");
foreach (var r in robots) {
Console.WriteLine("Robot id={0}, age={1}", r.id,r.age); //output. if this is web app you will likely do this differently
}

References
http://msdn.microsoft.com/en-us/magazine/ff796227.aspx